CN-122019641-A - Multi-time-scale prediction system based on meteorological grid

Abstract

The embodiment of the invention relates to a multi-time scale prediction system based on a meteorological grid, which comprises a data receiving module, a database updating module, a grid database, a queue pool updating module and a full grid queue Chi Hefu interface module.

Inventors

• PAN XUGUANG
• BAI GUODONG
• ZHANG XIAOFENG
• Huang Benfeng
• WU QIANG
• XU LIJUN
• Ha Yanli
• Meng Wentong
• YANG LIN
• ZHOU ZHIBO
• Yuan Haibao

Assignees

• 烟台市气象服务中心(烟台市专业气象台)
• 北京十环科技有限公司

Dates

Publication Date

20260512

Application Date

20240102

Claims (12)

1. 1. A multi-time scale prediction system based on a meteorological grid is characterized by comprising a data receiving module, a database updating module, a grid database, a queue pool updating module and a full grid queue Chi Hefu interface module; The system comprises a data receiving module, a database updating module, a grid database, a queue pool updating module, a service interface module and a network interface module, wherein the data receiving module is connected with an external weather prediction center and also connected with the database updating module; the data receiving module is used for periodically acquiring the weather forecast data of all weather grids in the whole period within 24 multiplied by 7 hours in the future from the weather forecast center according to the preset data acquisition frequency, and sending the weather forecast data to the database updating module as a corresponding first forecast data set; the database updating module is used for carrying out data updating processing on all first grid data lists of the grid database according to the first prediction data set and the first grid list of the grid database, and sending a queue updating instruction to the queue pool updating module when the data updating processing is finished; the grid database is used for storing the first grid list and a plurality of first grid data lists; the queue pool updating module is used for carrying out queue updating processing on all time scale queues of all first grid queue pools in the full grid queue pool according to the grid database when the queue updating instruction is received; The full-grid queue pool is used for storing a plurality of first grid queue pools, the first grid queue pools are in one-to-one correspondence with the meteorological grids, each first grid queue pool comprises a total number N of queues and N ith time scale queues, a queue index i is not less than 1 and not more than N, each ith time scale queue corresponds to a time scale s i ,1≤s i which takes an hour as a unit and is not more than 24, s i can be divided by 24, each ith time scale queue consists of a total number M i first queue records, M i ＝(24×7)/s i , each ith time scale queue is used for initializing the internal M i first queue records into empty records with all field contents being empty when the queues are initialized, and each ith time scale queue is also used for managing the queue records according to a first-in-first-out principle of a circular queue when one first queue record is added each time; The service interface module is used for receiving a first service application sent by the client, and performing weather forecast service data preparation processing according to the first service application, the grid database and the full grid queue pool to obtain corresponding first service feedback data and sending the corresponding first service feedback data back to the client.
2. 2. The weather grid based multi-time scale prediction system as claimed in claim 1, wherein, The first prediction data set comprises a plurality of first grid data subsets, wherein the first grid data subsets are in one-to-one correspondence with the meteorological grids, and the first grid data subsets comprise first grid identifications and first prediction data sequences; The first prediction data sequence is formed by sequencing 24 multiplied by 7 first prediction data records in time sequence, the first prediction data records comprise a first future time period, first weather phenomenon prediction data, first precipitation prediction data, first humidity prediction data, first average wind power prediction data, first gust wind power prediction data, first wind direction prediction data, first highest temperature prediction data, first lowest temperature prediction data, first minimum visibility prediction data and first cloud prediction data, the first future time period is composed of a pair of starting time and ending time, time information of the starting time and the ending time of the first future time period comprises year, month and day information and hour information, the time interval of the starting time and the ending time of the first future time period is one hour, in the first prediction data sequence, the year, month and day information of the starting time of the first future time period of the first prediction data records corresponds to the year, month and day information of the current time, the starting time of the first future period of the first predicted data record corresponds to the ending time of the first future period of the first predicted data record, the first weather phenomenon predicted data is a weather phenomenon phrase with one or more weather phenomenon names, the first precipitation predicted data is a precipitation value in millimeter, the first humidity predicted data is a humidity percentage, the first average wind power predicted data is a wind power level range consisting of a minimum wind power level and a maximum wind power level, the first gust wind power predicted data is a wind power level, the first wind direction predicted data is a wind direction name, the first highest temperature predicted data and the first lowest temperature predicted data are respectively a corresponding degree value, the first minimum visibility predicted data is a visibility value in kilometers, and the first cloud amount predicted data is a cloud amount percentage; the first grid list comprises a plurality of first grid records, wherein the first grid records are in one-to-one correspondence with the meteorological grids and also in one-to-one correspondence with the first grid data list, the first grid records comprise a first grid identification field, a first grid region name field, a first grid center point map coordinate field, a first grid vertex map coordinate set field and a first grid data list identification field, the first grid vertex map coordinate set field comprises a plurality of first grid vertex map coordinates, and the first grid data list identification field of each first grid record is matched with a first data list identification of the corresponding first grid data list; each first grid data list corresponds to a unique first data list identifier, the first grid data list comprises a plurality of first grid data records, each first grid data record comprises a first future period field, a first weather phenomenon field, a first precipitation field, a first humidity field, a first average wind force field, a first gust wind force field, a first wind direction field, a first highest temperature field, a first lowest temperature field, a first minimum visibility field and a first cloud amount field, each first future period field comprises a first starting time and a first ending time, time information of the first starting time and the first ending time comprises year and month information and hour information, a time interval from the first starting time to the first ending time is one hour, each first weather phenomenon field is a precipitation value with one or more weather phenomenon names, each first weather phenomenon field is a precipitation value in millimeters, each first humidity field is a first humidity value, each first minimum temperature field is a first peak wind force value, each first peak wind force value is a first peak wind force value, each first peak value is a first peak value, each first peak value is a peak value, and each peak value is a peak value; The first queue record of each ith time scale queue comprises a second future period field, a second weather phenomenon field, a second precipitation amount field, a second humidity field, a second average wind power field, a second gust wind power field, a second wind direction field, a second highest temperature field, a second lowest temperature field, a second minimum visibility field and a second cloud amount field, wherein the second future period field comprises a second starting time and a second ending time, time information of the second starting time and the second ending time comprises year, month and day information and hour information, and a time interval from the second starting time to the second ending time is matched with the corresponding time scale s i ; the first service application comprises a first service type, first application data and a first feedback data type, wherein the first service type comprises a full-network service type, a grid service type and a location service type, and the first feedback data type comprises a webpage file type, an image file type, a table file type and a text file type; The first application data comprises a first prediction starting time, a first prediction day and a first time scale when the first service type is a full-network service type, the first application data comprises a second prediction starting time, a second prediction day, a second time scale and a first prediction grid mark when the first service type is a grid service type, and the first application data comprises a third prediction starting time, a third prediction day, a third time scale and a first position coordinate when the first service type is a position service type.
3. 3. The weather grid based multi-time scale prediction system of claim 2, wherein, The database updating module is specifically configured to take each first grid data subset of the first prediction data set as a corresponding current grid data subset when data updating is performed on all first grid data lists of the grid database according to the first prediction data set and a first grid list of the grid database, take the first grid identification of the current grid data subset and the first prediction data sequence as corresponding current grid identification and a current prediction data sequence, take the first grid data list corresponding to the first grid data list identification field of the first grid record, in which the first grid identification field is matched with the current grid identification, as a corresponding current grid data list, perform data updating on the current grid data list according to the current prediction data sequence, and confirm that the data updating is finished when all the first grid data lists corresponding to all the first grid data subsets are completely data updated.
4. 4. The multi-time scale prediction system based on a meteorological grid of claim 3, The database updating module is specifically configured to traverse the first predicted data record of the current predicted data sequence when the current grid data list is updated according to the current predicted data sequence; and traversing, wherein the first prediction data record of the current traversal is used as a corresponding current prediction data record; and using the first grid data record in the current grid data list, in which the first future period field is matched with the first future period of the current predicted data record, as a corresponding current grid data record; and when the grid data record is empty, adding a new first grid data record to the current grid data list as the corresponding current grid data record and setting the first future period field of the current grid data record as the first future period of the current predicted data record, and setting the first weather phenomenon field, the first precipitation amount field, the first humidity field, the first average wind force field, the first gust wind force field, the first wind direction field, the first maximum temperature field, the first minimum visibility field and the first cloud amount field of the current grid data record as the corresponding first weather phenomenon predicted data, the first precipitation amount predicted data, the first humidity predicted data, the first average wind force predicted data, the first gust wind force predicted data, the first wind direction predicted data, the first maximum temperature predicted data, the first minimum temperature predicted data, the first temperature predicted data in the current predicted data record, the first minimum visibility prediction data and the first cloud cover prediction data; and when the traversal is finished, confirming that the current grid data list finishes data updating.
5. 5. The weather grid based multi-time scale prediction system of claim 2, wherein, The queue pool updating module is specifically configured to take each first grid queue pool of the full grid queue pool as a corresponding current grid queue pool when the queue updating process is performed on all time scale queues of all first grid queue pools in the full grid queue pool according to the grid database, take the first grid data list corresponding to the current grid queue pool in the grid database as a corresponding current grid data list, extract 24×7 first grid data records recently updated in the current grid data list, sort the first grid data records according to time sequence order to form a corresponding current grid data record sequence, perform queue updating on the current grid queue pool according to the current grid data record sequence, and confirm that the current queue updating process is finished when all the first grid queue pools complete the queue updating process.
6. 6. The grid-based multi-time scale prediction system of claim 5, The queue pool updating module is specifically configured to, when the current grid queue pool is updated according to the current grid data record sequence: step 61, taking the first ith time scale queue of the current grid queue pool as a corresponding current time scale queue; Step 62, taking the time scale s i corresponding to the current time scale queue as a corresponding current time scale s * , taking the total number M i of queue records corresponding to the current time scale queue as a corresponding current total number M * of queues, equally dividing the future 24×7 hours into third future time periods T j of the current total number M * of queues according to the current time scale s * , wherein the time scale index j is less than or equal to 1 and less than or equal to M * , and the time period length of each third future time period T j is consistent with the current time scale s * ; Step 63, extracting the first grid data records of which the first future time period field is in each third future time period T j in the current grid data record sequence to form a corresponding first record set G j , and performing data fusion processing on each first record set G j to obtain a corresponding first fusion record R j , wherein the number of the first grid data records of each first record set G j is consistent with the current time scale s * , the first fusion record R j comprises a first time period fusion field, a first atmospheric phenomenon fusion field, a first precipitation fusion field, a first humidity fusion field, a first average wind fusion field, a first gust wind fusion field, a first wind direction fusion field, a first highest temperature fusion field, a first lowest temperature fusion field, a first minimum visibility fusion field and a first cloud fusion field, and the first time period fusion field of the first fusion record R j is matched with the corresponding third future time period T j ; Step 64, traversing the third future period T j ; and traversing, taking the third future period T j of the current traversal as the corresponding current period; the first fusion record R j corresponding to the current time period is taken as a corresponding current fusion record, the first queue record with the second future time period field matched with the current time period in the current time scale queue is taken as the corresponding current queue record, whether the current queue record is empty or not is identified, if the current queue record is empty, the first time period fusion field, the first weather phenomenon fusion field, the first precipitation fusion field, the first humidity fusion field, the first average wind fusion field, the first gust wind fusion field, the first wind direction fusion field, the first highest temperature fusion field, the first lowest temperature fusion field, the first minimum visibility fusion field and the first cloud fusion field of the current time scale queue are taken as the corresponding second future time period field, the second weather phenomenon field, the second precipitation amount field, the second humidity, the second average wind phenomenon fusion field, the first average wind concentration field, the first gust wind concentration field, the first maximum temperature fusion field, the first minimum visibility fusion field and the first cloud amount fusion field of the current queue record are taken as the corresponding second future time period field, the second weather phenomenon field, the second cloud amount of the current queue record, the second weather phenomenon is added to the first time scale queue record, the second time scale record, the current time scale record and the second cloud record is added, resetting the second humidity field, the second average wind field, the second gust wind field, the second wind direction field, the second maximum temperature field, the second minimum visibility field and the second cloud amount field to the corresponding first atmospheric phenomenon fusion field, the first precipitation fusion field, the first humidity fusion field, the first average wind fusion field, the first gust wind fusion field, the first wind direction fusion field, the first maximum temperature fusion field, the first minimum visibility fusion field and the first cloud amount fusion field in the current fusion record; Step 65, after traversing, identifying whether the current time scale queue is the last ith time scale queue of the current grid queue pool, if not, taking the next ith time scale queue of the current grid queue pool as a new current time scale queue, and turning to step 62, if so, confirming that the current grid data recording sequence completes queue updating.
7. 7. The grid-based multi-time scale prediction system of claim 6, The queue pool updating module is specifically configured to use each first record set G j as a corresponding current record set when the data fusion processing is performed on each first record set G j to obtain a corresponding first fusion record R j ; The third future time period T j corresponding to the current record set is taken as a corresponding first set time period, the first precipitation amount fields of all the first grid data records of the current record set are summed to obtain a corresponding first precipitation amount sum, the first humidity fields of all the first grid data records of the current record set are averaged to obtain a corresponding first average humidity, the first average wind power field of all the first grid data records of the current record set is taken as a corresponding first set time period, the first minimum wind power level with the smallest value and the first maximum wind power level with the largest value are taken as a corresponding second wind power level range, the first gust wind power field with the largest wind power level in all the first grid data records of the current record set is taken as a corresponding second gust wind power level, the first average wind power field with the largest wind power level in all the first grid data records of the current record set is taken as a corresponding second gust wind power level, the first maximum wind power field in all the current record set is taken as a corresponding first wind power field of the current record set, and the first maximum wind power level in all the first grid data record set is taken as a corresponding first wind power field of the highest temperature, the method comprises the steps of obtaining a first grid data record of a current record set, obtaining a first cloud amount field of the current record set, obtaining a first minimum temperature field of the current record set, obtaining a corresponding first minimum temperature field of the first grid data record of the current record set, obtaining a corresponding first average cloud amount by means of average calculation of the first cloud amount fields of all the first grid data records of the current record set, obtaining a corresponding first phrase sequence by means of de-duplication and merging of weather phenomenon phrases of the first weather phenomenon fields of all the first grid data records of the current record set, and obtaining a corresponding first weather phenomenon phrase sequence by means of weather phenomenon phrase fusion processing according to a time scale s i corresponding to the current record set, a first precipitation amount sum, the first phrase sequence, the first minimum visibility and the first average cloud amount; And the obtained first aggregate period, the first weather phenomenon phrase, the first total precipitation amount, the first average humidity, the second wind level range, the second gust wind level, the first wind direction name, the first maximum temperature, the first minimum visibility, and the first average cloud amount are used as corresponding first period fusion fields, the first weather phenomenon fusion fields, the first precipitation amount fusion fields, the first humidity fusion fields, the first average wind fusion fields, the first gust wind fusion fields, the first wind direction fusion fields, the first maximum temperature fusion fields, the first minimum visibility fusion fields, and the first cloud amount fusion fields to form a corresponding first fusion record R j .
8. 8. The grid-based multi-time scale prediction system of claim 7, The queue pool updating module is specifically configured to, when the weather phenomenon phrase fusion process is performed according to the time scale s i corresponding to the current record set and the first precipitation sum, the first phrase sequence, the first minimum visibility, and the first average cloud amount to obtain a corresponding first weather phenomenon phrase, Inquiring a first preset corresponding relation table for reflecting the corresponding relation between a time scale and weather parameters, and extracting a first weather parameter set field of a first corresponding relation record, in which a first time scale field is matched with the time scale s i corresponding to the current record set, in the first corresponding relation table to serve as a corresponding first weather parameter set, wherein the first corresponding relation table comprises a plurality of first corresponding relation records, the first corresponding relation record comprises the first time scale field and the first weather parameter set field, and the first weather parameter set field comprises a gross rain volume range, a small rain volume range, a medium rain volume range, a heavy storm volume range, an extra heavy storm volume range, a fine cloud volume range, a cloudy cloud volume range, a haze visibility range, a light haze visibility range and a fog visibility range; when the sum of the first precipitation amounts is 0, carrying out weather phenomenon phrase fusion according to the first weather parameter set, the first average cloud amount and the first minimum visibility to obtain the corresponding first weather phenomenon phrase; and when the sum of the first precipitation amounts is greater than 0, carrying out weather phenomenon phrase fusion according to the first weather parameter set, the first precipitation amount and the first phrase sequence to obtain the corresponding first weather phenomenon phrase.
9. 9. The weather grid based multi-time scale prediction system as claimed in claim 8, wherein, The queue pool updating module is specifically configured to, when the weather phenomenon phrase fusion is performed according to the first weather parameter set, the first average cloud amount, and the first minimum visibility to obtain the corresponding first weather phenomenon phrase, The first average cloud amount is identified, if the first average cloud amount meets the sunny cloud amount range of the first weather parameter set, a corresponding first weather phenomenon keyword is set as 'sunny', if the first average cloud amount meets the cloudy cloud amount range of the first weather parameter set, a corresponding first weather phenomenon keyword is set as 'cloudy', and if the first average cloud amount meets the cloudy cloud amount range of the first weather parameter set, a corresponding first weather phenomenon keyword is set as 'cloudy'; The first minimum visibility is identified, if the first minimum visibility meets the haze-free visibility range of the first weather parameter set, a corresponding second weather phenomenon keyword is set to be empty, if the first minimum visibility meets the light haze visibility range of the first weather parameter set, a corresponding second weather phenomenon keyword is set to be light haze, if the first minimum visibility meets the medium haze visibility range of the first weather parameter set, a corresponding second weather phenomenon keyword is set to be fog, and if the first minimum visibility meets the thick haze visibility range of the first weather parameter set, a corresponding second weather phenomenon keyword is set to be thick fog; And identifying whether the obtained second weather phenomenon keywords are empty, if so, setting the corresponding first weather phenomenon phrases as the first weather phenomenon keywords, and if not, setting the corresponding first weather phenomenon phrases as phrases formed by sequentially splicing the first weather phenomenon keywords, preset isolation symbols and the second weather phenomenon keywords, wherein the preset isolation symbols default to commas.
10. 10. The weather grid based multi-time scale prediction system as claimed in claim 8, wherein, The queue pool updating module is specifically configured to, when the weather phenomenon phrase fusion is performed according to the first weather parameter set, the first precipitation amount, and the first phrase sequence to obtain the corresponding first weather phenomenon phrase, Setting the corresponding first weather phenomenon phrase as "hair rain" when the first precipitation amount sum meets the hair rain amount range of the first weather parameter set; When the first precipitation sum meets the small rain amount range of the first weather parameter set, identifying keywords in the first phrase sequence, setting the corresponding first weather phenomenon phrase as small rain if the first phrase sequence does not contain keywords of thunder and hail, setting the corresponding first weather phenomenon phrase as thunder and hail if the first phrase sequence contains keywords of thunder and hail, and setting the corresponding first weather phenomenon phrase as thunder and hail if the first phrase sequence contains keywords of thunder and hail and setting the corresponding first weather phenomenon phrase as thunder and hail; When the first precipitation amount sum meets the medium rainfall amount range, the heavy rainfall amount range or the extra heavy rainfall amount range of the first air parameter set, identifying keywords in the first phrase sequence; if the first phrase sequence does not contain the keyword 'thunder' and the keyword 'hail', setting the corresponding first atmospheric phenomenon phrases as corresponding 'medium rain', 'heavy storm' or 'extra heavy storm'; setting the corresponding first weather phenomenon phrases as corresponding medium rain, thunder and hail, heavy storm, thunder and hail, or super heavy storm, and thunder and hail, if the first phrase sequence contains the keyword thunder and hail, and does not contain the keyword hail, setting the corresponding first weather phenomenon phrases as corresponding medium rain, thunder and heavy rain, thunder and storm, and thunder, heavy storm, and thunder.
11. 11. The weather grid based multi-time scale prediction system of claim 2, wherein, The ith time scale queue is specifically configured to, when performing queue record management according to a first-in-first-out rule of a circular queue when one first queue record is added each time, regarding the first queue record added at the time as a corresponding current added record when one first queue record is newly added each time, counting the number of empty records with all field contents being empty in the current ith time scale queue to obtain a corresponding current empty record total number Q, identifying the current empty record total number Q, regarding the (M i -Q+1) th first queue record in the current ith time scale queue as a corresponding current modified record if the current empty record total number Q is greater than 0, regarding the first queue record with the earliest time of the second future period field in the current ith time scale queue as the corresponding current modified record if the current empty record total number Q is equal to 0, and setting each corresponding field in the current modified record based on each field of the current added records.
12. 12. The weather grid based multi-time scale prediction system of claim 2, wherein, The service interface module is specifically configured to extract the corresponding first service type, the first application data, and the first feedback data type from the first service application when the weather forecast service data is prepared according to the first service application, the grid database, and the full grid queue pool to obtain corresponding first service feedback data and send the corresponding first service feedback data back to the client; identifying the first service type; If the first service type is a full-network service type, extracting the corresponding first prediction starting time, the first prediction days and the first time scale from the first application data, adding the first prediction starting time and the first prediction days to obtain the corresponding first prediction ending time, forming a corresponding first prediction time period by the first prediction starting time and the first prediction ending time, using the ith time scale queue matched with the first time scale by the time scale s i in each first grid queue pool of the full-grid queue pool as a corresponding first grid matching queue, extracting the first queue records with intersections of the second future time period fields and the first prediction time period in each first grid matching queue, sequentially arranging the first grid records according to time to form a corresponding first grid prediction list, and forming a corresponding first prediction set by all the obtained first grid prediction lists; If the first service type is a grid service type, extracting the corresponding second prediction starting time, the second prediction days, the second time scale and the first prediction grid identification from the first application data, adding the second prediction starting time and the second prediction days to obtain the corresponding second prediction ending time, forming a corresponding second prediction time period by the second prediction starting time and the second prediction ending time, extracting the first grid queue pool corresponding to the first prediction grid identification from the full grid queue pool to serve as a corresponding first matching grid pool, taking the ith time scale queue matched by the time scale s i and the second time scale in the first matching grid pool to serve as a corresponding second grid matching queue, extracting the first queue record with intersection of the second future time period field and the second prediction time period in the second grid matching queue to form a corresponding first grid list in time sequence, and forming the first prediction list by the first grid list; If the first service type is a location service type, extracting the corresponding third predicted starting time, third predicted days, third time scale and first location coordinates from the first application data; the method comprises the steps of obtaining a first prediction starting time, obtaining a first position coordinate of a first grid list, obtaining a corresponding first linear distance by adding the first prediction starting time and the first prediction days to obtain a corresponding third prediction ending time, obtaining a corresponding third prediction period by forming a corresponding third prediction time by the third prediction starting time and the third prediction ending time, obtaining corresponding first recognition results by recognizing the inclusion and non-inclusion relation between a polygon map area formed by the first grid vertex map coordinate set field of each first grid record of the first grid database and the first position coordinate, obtaining a corresponding first linear distance by calculating a linear distance between the first grid center point map coordinate field of each first grid record and the first position coordinate, taking the first grid record corresponding to the first recognition result in particular to the inclusion relation as a corresponding first to be used as a corresponding first to be selected grid record, taking the first to be used as a corresponding to be selected grid record, taking the first to be matched grid record corresponding to the first position coordinate of the first grid list, taking the first grid list corresponding to the first position coordinate list corresponding to the first grid list corresponding to the first position coordinate list, taking the first grid list corresponding to the first grid vertex map area surrounded by the first grid vertex map coordinate set field of each first grid vertex map of the first to be used as a corresponding first grid list corresponding to be matched with the first grid list corresponding to a first position list corresponding to be corresponding to a first position coordinate list The first queue record of the intersection of the second future period field and the third prediction period in the third grid matching queue is extracted to be arranged in time sequence to form a corresponding first grid prediction list, and the obtained first grid prediction list forms a corresponding first grid prediction list set; Identifying the first feedback data type; if the first feedback data type is a webpage file type, all the first grid prediction columns of the obtained first grid prediction list set are put into a preset webpage template to be subjected to webpage file conversion processing to obtain a corresponding first webpage file, and the obtained first webpage file is used as the corresponding first service feedback data; if the first feedback data type is an image file type, performing table image conversion processing on each first grid prediction list of the obtained first grid prediction list set to obtain a corresponding first table image, forming a corresponding first image file by all the obtained first table images, and taking the obtained first image file as the corresponding first service feedback data; if the first feedback data type is a table file type, performing electronic table conversion processing on each first grid prediction list of the obtained first grid prediction list set to obtain a corresponding first electronic table, forming a corresponding first table file by all the obtained first electronic tables, and taking the obtained first table file as corresponding first service feedback data; and sending the obtained first service feedback data back to the client.

Description

Multi-time-scale prediction system based on meteorological grid Technical Field The invention relates to the technical field of data processing, in particular to a multi-time-scale prediction system based on a meteorological grid. Background The weather grid forecast (or prediction) is also called weather intelligent grid forecast (or prediction), and specifically refers to dividing a map into a plurality of unit grids according to a certain grid precision, wherein each unit grid is a weather grid, and weather forecast processing is carried out by taking the weather grid as a unit. With the deep development of weather grid prediction technology, most weather prediction centers can predict weather information of each grid in a future 24×7 period. However, in practical application, we find some problems that 1) the time scale of grid prediction data with 24×7 specifications, such as a weather prediction center, is too single, only supports a single time scale, specifically 1 hour, and cannot meet the multi-scale requirements of different application scenarios, for example, the time scale required in an application scenario of personal travel is conventionally 1 hour, the time scale required in an application scenario of open-air production construction is conventionally 1 or 2 hours, the scale requirements required in an application scenario of environment monitoring are typically 3, 4, 6, 8 or 12 hours, and the like, and 2) most weather prediction centers do not support complex personalized query modes, such as query according to grid identification+time scale, query according to positioning coordinates (or positions) +time scale, and the like. Disclosure of Invention The invention aims to overcome the defects of the prior art and provides a multi-time scale prediction system based on a meteorological grid, which comprises a data receiving module, a database updating module, a grid database, a queue pool updating module and a full-grid queue Chi Hefu service interface module, wherein the data receiving module is used for acquiring weather forecast data of all meteorological grids within 24×7 hours in the future from a meteorological prediction center as a prediction data set and sending the prediction data set to the database updating module, the database updating module is used for updating the data of the grid database according to the prediction data set and sending a queue updating instruction to the queue pool updating module when the updating is finished, the queue pool updating module is used for carrying out queue updating on all time scale queues of all first grid queue pools in the full-grid queue pool according to the grid database when the queue updating instruction is received, and the service interface module is used for receiving a first service application sent by an external optional client and carrying out meteorological prediction service data preparation according to the first service application, the grid database and the full-grid queue pool and obtaining corresponding service feedback data to be sent back to the client. The full-grid queue pool can create a plurality of time-scale predicted data queues for each meteorological grid, the queue pool updating module can timely refresh each time-scale queue of each meteorological grid when receiving the latest predicted data set, and the service interface module can provide full-grid query service, single-grid query service based on grid identification and nearest-grid query service based on position, provide time-scale parameters in each query service to meet multi-scale query, and provide various feedback forms such as web pages, images, tables, texts and the like when the query result is fed back. The method and the system can solve the problem that the time scale of the grid prediction data output by the weather prediction center is too single, and can also improve the query richness of the grid prediction data. In order to achieve the above purpose, the embodiment of the invention provides a multi-time scale prediction system based on a meteorological grid, which comprises a data receiving module, a database updating module, a grid database, a queue pool updating module and a full grid queue Chi Hefu interface module; The system comprises a data receiving module, a database updating module, a grid database, a queue pool updating module, a service interface module and a network interface module, wherein the data receiving module is connected with an external weather prediction center and also connected with the database updating module; the data receiving module is used for periodically acquiring the weather forecast data of all weather grids in the whole period within 24 multiplied by 7 hours in the future from the weather forecast center according to the preset data acquisition frequency, and sending the weather forecast data to the database updating module as a corresponding first forecast data set; the database updating module is us