CN-121980306-A - Method for realizing fluoride deep treatment and recycling recovery

Abstract

The invention discloses a method for realizing fluoride deep treatment and recycling, which particularly relates to the technical field of fluoride deep treatment and recycling, and comprises the steps of acquiring process data of a fluoride deep treatment device and a recycling device in an operation process, continuously collecting the process data, outputting an original monitoring data sequence arranged in time sequence, carrying out time window segmentation processing on the continuously collected process data, firstly identifying a corresponding operation stage identifier, then carrying out semantic judgment on a monitoring signal under stage constraint, and optimizing a driving strategy based on deviation between a semantic result and a treatment target or a recycling target, thereby avoiding the problems of reaction state judgment, process switching and recycling path selection distortion caused by understanding the monitoring data in a fixed interpretation mode.

Inventors

• SHI ZHENFENG
• LI BAOLU
• XI ZHIPENG
• WEI WENXIN
• YU QUN
• WANG WENYUE
• Gao Longqun
• HOU JIANGUO
• ZHAO NING
• LI ZHENG
• SHANG YONG
• QI CUIHONG
• CHEN ZHANGHAO
• WANG XIAOHONG
• LI SHUZHI
• SHI MIN
• ZHANG XIAOLIN

Assignees

• 山东环瑞生态科技有限公司

Dates

Publication Date

20260505

Application Date

20260115

Claims (9)

1. 1. A method for effecting fluoride depth remediation and recycling, comprising: S1, acquiring process data of a fluoride deep treatment device and a recycling recovery device in the operation process, continuously acquiring the process data, and outputting an original monitoring data sequence arranged in time sequence; S2, removing obvious noise in an original monitoring data sequence, carrying out segmentation processing on the monitoring data according to preset time windows, and comparing the segmentation processing result with preset stage division conditions to identify characteristic data for stage judgment in each time window; s3, repeatedly identifying corresponding characteristic data in a plurality of sections of adjacent time windows, and outputting corresponding stage identifiers when the characteristic data in the adjacent time windows meet corresponding stage division thresholds; s4, extracting feature information for semantic judgment from the process data based on the current stage identification, comparing the extracted feature information with preset semantic judgment conditions item by item, and outputting a corresponding semantic result according to the comparison result; s5, comparing the semantic result with a preset fluoride deep treatment target or a recycling recovery target, selecting or combining a control strategy from a strategy library, outputting a strategy parameter set comprising deep treatment strategy parameters or recycling recovery strategy parameters, and controlling and outputting a deep treatment process or recycling recovery process according to the strategy parameter set.
2. 2. A method for achieving fluoride depth remediation and recycling according to claim 1, wherein: in S1, setting an independent acquisition channel for process data, and enabling various process data to enter a data acquisition flow in an independent acquisition channel mode, wherein the process data comprise six types of data, namely raw water fluoride ion concentration data, deep treatment effluent fluoride ion concentration data, suspended particle size distribution data, stirring motor current data, sedimentation liquid level height data and recycling mother liquor fluoride ion concentration data; Performing periodic sampling on independent acquisition channels corresponding to the six types of process data in a uniform sampling period, respectively acquiring the six types of process data in each sampling period, and attaching a corresponding sampling time mark to each type of process data to obtain process data with a sampling time mark; synchronous combination is carried out on various process data acquired under the same sampling time mark, and the process data are written into the same monitoring record according to a preset sequence to form a monitoring data point containing six types of process data; and sequentially arranging the monitoring data points obtained in a plurality of continuous sampling periods according to the sequence of the sampling time marks, and connecting the sequentially arranged monitoring data points to form an original monitoring data sequence covering the running process of the fluoride deep treatment device and the recycling device.
3. 3. A method for achieving fluoride depth remediation and recycling according to claim 2, wherein: In S2, analyzing each monitoring data point in the original monitoring data sequence one by one, checking whether sampling values of six types of process data in the corresponding monitoring data points fall into a corresponding physical allowable range and a corresponding measuring range respectively, and if any sampling value exceeds the corresponding range, marking the process data of the corresponding type as obvious noise and eliminating the obvious noise; after eliminating the process data of the corresponding category marked as obvious noise, checking whether each monitoring data point still contains sampling values corresponding to the six categories of process data, and if the sampling value of any category is missing, eliminating the corresponding monitoring data point integrally to form a monitoring data sequence; Segmenting a monitoring data sequence according to the sequence of sampling time marks according to the preset time window size, dividing a monitoring data point with sampling time falling into the same time window range into a time window section, and distributing a corresponding window identifier for each time window section; In each time window section, taking the first monitoring data point in the time window section which is sequenced according to sampling time as a starting point, taking the last monitoring data point in the time window section as an ending point, respectively calculating sampling value differences between the starting point and the ending point of the six types of process data, determining corresponding change trends according to the magnitude relation of the differences of the process data, recording as an ascending trend when the difference is larger than zero, recording as a constant trend when the difference is equal to zero, and recording as a descending trend when the difference is smaller than zero; In the same time window, calculating the difference between the upper limit value and the lower limit value of all sampling values of the current data of the stirring motor to obtain the current fluctuation amplitude; And the difference value, the corresponding change trend and the current fluctuation amplitude of the six types of process data jointly form a candidate characteristic data set corresponding to the time window section, wherein the candidate characteristic data set comprises a difference value characteristic, a change trend characteristic and a current fluctuation amplitude characteristic.
4. 4. A method for achieving fluoride depth remediation and recycling recovery according to claim 3, wherein: s2, screening difference value features, change trend features and current fluctuation amplitude features in the candidate feature data set with threshold conditions preset for a deep treatment reaction stage, a deep treatment and recycling transition stage and a recycling separation stage in a stage division condition in each time window segment, wherein: For the difference features, respectively comparing the difference values of the six types of process data with an upper limit value and a lower limit value of a difference threshold value interval preset for each stage in the stage division condition, when the numerical value of a certain difference feature falls into the difference threshold value interval corresponding to any stage, recording the difference feature corresponding to the difference feature as the difference feature meeting the stage division condition, otherwise, removing the corresponding difference feature from the candidate feature data set; Aiming at the change trend characteristics, respectively comparing the change trends corresponding to the six types of process data with a preset allowable change trend set for each stage in the stage division conditions, when a certain change trend characteristic belongs to the allowable change trend set of any stage, recording the change trend characteristic corresponding to the change trend characteristic as the change trend characteristic meeting the stage division conditions, otherwise, removing the corresponding change trend characteristic from the candidate characteristic data set; Aiming at the current fluctuation amplitude characteristics, comparing the current fluctuation amplitude with an upper limit value and a lower limit value of a current fluctuation amplitude threshold value interval preset for each stage in the stage division condition, recording the current fluctuation characteristics of which the corresponding current fluctuation amplitude characteristics meet the stage division condition when the value of the current fluctuation amplitude falls into the current fluctuation amplitude threshold value interval corresponding to any stage, otherwise, eliminating the corresponding current fluctuation amplitude characteristics from the candidate characteristic data set; after the screening of the difference value characteristics, the change trend characteristics and the current fluctuation amplitude characteristics is completed, the difference value characteristics, the change trend characteristics and the current fluctuation amplitude characteristics which are reserved in the screening are used as characteristic data for stage judgment in the corresponding time window sections.
5. 5. The method for achieving fluoride depth remediation and recycling of claim 4, wherein: in S3, sequentially reading characteristic data for stage judgment corresponding to each time window section according to the sequence of the time window identifications, sequentially arranging all the time window sections to form a window processing sequence, and setting a preset initial stage identification for the first time window section in the window processing sequence as a previous stage identification; When a current time window segment in a window processing sequence is processed, classifying difference features, change trend features and current fluctuation amplitude features in feature data of which the current time window segment is used for stage judgment respectively to form a difference feature set, a change trend feature set and a current fluctuation amplitude feature set; Aiming at each stage of the deep treatment reaction stage, the deep treatment and recycling transition stage and the recycling separation stage, respectively counting the difference characteristic quantity of a difference value threshold interval corresponding to the corresponding stage in a difference characteristic set of a current time window section, counting the change trend characteristic quantity of a change trend allowed to be in a change trend characteristic set belonging to the corresponding stage, counting the current fluctuation amplitude characteristic quantity of a current fluctuation amplitude threshold interval corresponding to the corresponding stage in a current fluctuation amplitude characteristic set, and summing the characteristic quantity meeting the conditions of the difference value threshold interval, the allowed change trend set and the current fluctuation amplitude threshold interval to obtain the stage accumulated characteristic quantity of the corresponding stage; Selecting a stage with the stage accumulation feature number at the upper limit value from the stage accumulation feature numbers of the three stages as a candidate stage of the current time window section, and determining a stage identifier corresponding to the candidate stage at the upper limit value as a stage identifier of the current time window section when the stage accumulation feature number at the upper limit value is greater than zero and only one stage corresponds to the maximum stage accumulation feature number; When the phase accumulation feature numbers of the three phases are zero or two or more phases are corresponding to the same phase accumulation feature number at the upper limit value, the phase switching is not carried out on the current time window segment, and the previous phase identifier is directly used as the phase identifier of the current time window segment; writing the phase identification of the current time window into a phase identification sequence, updating the phase identification of the current time window into the previous phase identification of the next time window, and repeatedly executing the phase identification process of the corresponding time window for the rest time windows in the window processing sequence until all the time windows finish the determination of the phase identification.
6. 6. The method for achieving fluoride depth remediation and recycling of claim 5, wherein: In S4, after determining a stage identifier corresponding to the current time window segment, reading monitoring data points in the current time window segment, determining a start monitoring data point and an end monitoring data point according to the sequence of sampling time marks, and respectively extracting sampling values corresponding to the depth treatment effluent fluoride ion concentration data, the suspended particle size distribution data, the sedimentation liquid level height data and the recycling recovery mother liquor fluoride ion concentration data in the start monitoring data point and the end monitoring data point; calculating a fluoride ion concentration variation, a concentration variation rate per unit time, a suspended particle size variation, a particle size variation trend, a sedimentation velocity and a fluoride ion concentration variation of the recycling recovery mother liquor based on a sampling value between the initial monitoring data point and the end monitoring data point, wherein: the fluoride ion concentration variation is the difference value of the fluoride ion concentration sampling value of the deep treatment effluent in the end monitoring data point and the initial monitoring data point; The concentration change rate in unit time is the ratio of the fluorine ion concentration change amount to the sampling time difference of the two monitoring data points; The particle size variation of the suspended particles is the difference value of particle size statistics of the end monitoring data point and the start monitoring data point; The particle size change trend is determined as an ascending trend, a constant trend or a descending trend according to the sign of the particle size change quantity of the suspended particles, wherein the ascending trend is determined when the particle size change quantity of the suspended particles is larger than zero, the constant trend is determined when the particle size change quantity of the suspended particles is equal to zero, and the descending trend is determined when the particle size change quantity of the suspended particles is smaller than zero; the sedimentation speed is the ratio of the liquid level difference value of the initial monitoring data point and the final monitoring data point to the sampling time difference; the change amount of the concentration of the fluorine ions in the mother liquor recycled is the difference value of the concentration of the fluorine ions in the mother liquor in the initial monitoring data point and the ending monitoring data point; The fluorine ion concentration variation, the concentration variation rate in unit time, the particle size variation of suspended particles, the sedimentation velocity and the fluorine ion concentration variation of the recycling mother liquor are taken as numerical characteristics, and the particle size variation trend is taken as trend characteristics.
7. 7. The method for achieving fluoride depth remediation and recycling of claim 6, wherein: S4, when the current stage identification belongs to the deep treatment reaction stage, based on the fluorine ion concentration variation, the concentration variation rate in unit time, the suspended particle size variation and the particle size variation trend, comparing the corresponding numerical type features and trend type features with preset semantic type judgment conditions item by item, wherein each semantic type corresponds to at least one semantic judgment sub-condition set formed by a numerical threshold zone or a variation trend set; When the current stage identification belongs to a recycling recovery separation stage, based on the sedimentation velocity and the change quantity of the concentration of fluorine ions in the recycling recovery mother liquor, comparing the corresponding numerical value characteristics with preset semantic type judgment conditions item by item, and when all semantic judgment conditions corresponding to any semantic type are met, recording a separation stage semantic result with the corresponding semantic type being a current time window section; when the current stage identification belongs to the advanced treatment and recycling transition stage, the semantic type judgment is not executed, and a null semantic result set is directly written into the semantic result sequence; and sequencing the semantic results corresponding to each time window according to the time sequence of the time window to obtain a semantic result sequence arranged in time sequence.
8. 8. The method for achieving fluoride depth remediation and recycling of claim 7, wherein: In S5, after determining a stage identifier and a semantic result of a current time window segment, when the stage identifier belongs to a deep treatment reaction stage, reading a sampling value of the concentration of the deep treatment effluent fluoride ions from a monitoring data point corresponding to the current time window segment, comparing the sampling value with a target concentration interval set for the concentration of the effluent fluoride ions in a preset fluoride deep treatment target, recording the deviation of the effluent fluoride ions as a forward deviation when the concentration of the effluent fluoride ions is higher than the upper limit value of the target concentration interval, recording the deviation of the effluent fluoride ions as a negative deviation when the concentration of the effluent fluoride ions is lower than the lower limit value of the target concentration interval, and recording the deviation of the effluent fluoride ions as a zero deviation when the concentration of the effluent fluoride ions falls into the target concentration interval; When the current stage mark belongs to the recycling separation stage, reading the sedimentation velocity and the fluorine ion concentration variation of the recycling mother liquor from a monitoring data point corresponding to a current time window section, respectively comparing the sedimentation velocity with a corresponding target section in a preset recycling target, recording the sedimentation velocity deviation as a negative deviation when the sedimentation velocity is lower than a lower limit value of the target section, recording the sedimentation velocity deviation as a positive deviation when the sedimentation velocity is higher than an upper limit value of the target section, recording the sedimentation velocity deviation as a zero deviation when the sedimentation velocity falls into the target section, recording the fluorine ion concentration variation of the mother liquor as a negative deviation when the fluorine ion concentration variation of the mother liquor is lower than the lower limit value of the target section, recording the fluorine ion concentration variation of the mother liquor as a positive deviation when the fluorine ion concentration variation of the mother liquor is higher than the upper limit value of the target section, and recording the fluorine ion concentration variation of the mother liquor as a zero deviation when the fluorine ion concentration variation of the mother liquor falls into the target section; Determining a strategy optimization requirement of a current time window section based on the direction marks of the effluent concentration deviation, the sedimentation velocity deviation and the mother liquor concentration variation deviation, generating the strategy optimization requirement according to the following rule aiming at each direction mark, and recording the deviation index type corresponding to the direction mark as a current deviation index: when the water outlet concentration deviation is forward deviation, recording the strategy optimization requirement as an enhancement depth treatment function, and recording the deviation index type as a water outlet concentration deviation index; When the water outlet concentration deviation is negative deviation, recording the strategy optimization requirement as weakening the depth treatment effect, and recording the deviation index type as a water outlet concentration deviation index; When the sedimentation velocity deviation is negative deviation, recording the strategy optimization requirement as an enhanced sedimentation condition, and recording the deviation index type as a sedimentation velocity deviation index; when the sedimentation velocity deviation is forward deviation, recording the strategy optimization requirement as a settlement weakening condition, and recording the deviation index type as a sedimentation velocity deviation index; When the concentration variation deviation of the mother liquor is negative deviation, recording the strategy optimization requirement as the enhanced recycling capability, and recording the deviation index type as the concentration variation deviation index of the mother liquor; when the concentration variation deviation of the mother liquor is a forward deviation, recording the strategy optimization requirement as weakening the recycling capability, and recording the deviation index type as a concentration variation deviation index of the mother liquor; when any deviation index is zero deviation, a strategy optimization requirement is not generated for the corresponding deviation index.
9. 9. The method for achieving fluoride depth remediation and recycling of claim 8, wherein: The method comprises the steps of S5, searching a strategy optimization record matched with a corresponding strategy optimization demand from a pre-constructed strategy library according to each strategy optimization demand in the strategy optimization demands, wherein the strategy optimization record at least comprises a deviation index type, an applicable stage mark, a strategy optimization category, a parameter name, a parameter adjustment direction and a parameter adjustment step length; The deviation index type is used for representing which deviation source the strategy optimization record is applicable to, and comprises at least one of a water outlet concentration deviation index, a sedimentation speed deviation index and a mother liquor concentration change deviation index; The applicable stage identifier is used for indicating an applicable operation stage of the corresponding strategy optimization record, and comprises at least one of a deep treatment reaction stage or a recycling recovery separation stage; the policy optimization category is used for indicating which policy optimization requirements the corresponding policy optimization record is to respond to, including at least one of enhancing depth governance, weakening depth governance, enhancing settling conditions, weakening settling conditions, enhancing recycling capacity, or weakening recycling capacity; The parameter name is used for indicating the specific control parameters to be adjusted in the strategy optimization record, and the specific control parameters comprise at least one of dosage, dosage time interval, stirring rotating speed, sedimentation residence time, starting and stopping time of a recycling recovery unit, recovery flow, crystal classification mode and mother liquor discharge and reflux distribution proportion; The parameter adjustment direction is used to identify a specific adjustment direction for performing an increase, decrease, shorten, lengthen, raise, decrease, advance, or retard on the parameter; The parameter adjustment step length is used for giving the adjustment amplitude of the corresponding parameter when each time is adjusted; In the retrieval process, when the type of the deviation index in a certain strategy optimization record is consistent with the current deviation index, the applicable stage identifier is consistent with the current stage identifier, and the strategy optimization type is consistent with the current strategy optimization requirement, adding the corresponding strategy optimization record into a strategy candidate set; Reading each policy optimization record in the policy candidate set, and sequentially executing parameter adjustment operation according to the arrangement sequence of the policy optimization records in the policy candidate set, and executing one-time deterministic adjustment processing on the appointed parameter according to the parameter name, the parameter adjustment direction and the parameter adjustment step length in each policy optimization record, wherein the method specifically comprises the following steps: When the parameter name is the dosing quantity and the parameter adjustment direction is the decrease, subtracting the current dosing quantity from the parameter adjustment step length to obtain a new dosing quantity; When the parameter name is the dosing time interval and the parameter adjustment direction is prolonged, adding the current dosing time interval and the parameter adjustment step length to obtain a new dosing time interval; When the parameter name is stirring rotation speed and the parameter adjustment direction is decreasing, subtracting the current stirring rotation speed from the parameter adjustment step length to obtain a new stirring rotation speed; when the parameter name is settling residence time and the parameter adjustment direction is shortened, subtracting the current settling residence time from the parameter adjustment step length to obtain new settling residence time; When the parameter name is the recycling recovery unit start-stop time and the parameter adjustment direction is delay, the current start-stop time and the parameter adjustment step length are added to obtain a new start-stop time; when the parameter name is the recovery flow and the parameter adjustment direction is reduced, subtracting the current recovery flow from the parameter adjustment step length to obtain a new recovery flow; when the parameter name is a crystal grading mode, moving forwards or backwards by one grade according to the index sequence in a preset grading mode sequence according to the parameter adjustment direction, and taking the moved grading mode as a new crystal grading mode; when the parameter name is the mother liquor discharge and reflux distribution ratio and the parameter adjustment direction is the reflux ratio increasing, the current reflux ratio and the parameter adjustment step length are added and limited in an allowed interval to obtain a new reflux ratio, and the discharge ratio is updated to be a new reflux ratio subtracting; After finishing the adjustment of the corresponding parameters of the single strategy optimization record, carrying out item-by-item comparison on the adjusted parameter value and the corresponding parameter allowed value interval, setting the adjusted parameter value as the upper limit value of the corresponding parameter allowed value interval when the adjusted parameter value is larger than the upper limit value of the corresponding parameter allowed value interval, and setting the adjusted parameter value as the lower limit value of the corresponding parameter allowed value interval when the adjusted parameter value is smaller than the lower limit value of the corresponding parameter allowed value interval; When a plurality of strategy optimization records point to the same parameter in the strategy candidate set, absolute value operation is carried out on the deviation value corresponding to each strategy optimization record to obtain a deviation absolute value, the magnitudes of all the deviation absolute values are compared, the strategy optimization record with the deviation absolute value being at the upper limit value is selected as the strategy optimization record which is currently and uniquely executed, parameter adjustment operation is carried out on the parameter corresponding to the currently executed strategy optimization record, and other strategy optimization records pointing to the same parameter are not executed.

Description

Method for realizing fluoride deep treatment and recycling recovery Technical Field The invention relates to the technical field of fluoride deep treatment and recycling recovery, in particular to a method for realizing fluoride deep treatment and recycling recovery. Background In the existing fluoride deep treatment and recycling process, the monitoring data are generally directly regarded as physical and chemical indexes reflecting the real state of the system, and the control system judges whether the reaction process, sedimentation behavior, crystal growth trend and recycling conditions are met or not according to the data; however, in a continuous operation scenario, the same type of monitoring signal often corresponds to completely different practical meanings in different reaction contexts, for example, the decrease of the fluoride ion concentration may be derived from normal precipitation reaction, or may be derived from pseudo decrease caused by adsorption residue or complexing agent interference; Because the traditional control system cannot identify the semantic differences with similar surfaces but different essence, the traditional control system can misconsider the data change as stable and unified materialization process indication, so that systematic deviation occurs in reaction state judgment, process switching opportunity judgment and recovery path selection; it follows that a core problem is that the current technology lacks the ability to identify semantic changes in the monitored signal in different reaction phases, different chemical environments and different process contexts, yet understands the data in a fixed interpretation manner, resulting in a distortion of the policy decision basis. Disclosure of Invention In order to overcome the above-mentioned drawbacks of the prior art, an embodiment of the present invention provides a method for implementing advanced fluoride treatment and recycling, by performing time window segmentation processing on continuously collected process data, identifying corresponding operation stage identifiers first, then performing semantic judgment on a monitoring signal under stage constraint, and driving policy optimization based on deviation between a semantic result and a treatment target or recycling target, thereby avoiding problems of reaction state judgment, process switching and recycling path selection distortion caused by understanding the monitoring data in a fixed interpretation manner. In order to achieve the purpose, the invention provides the following technical scheme that the method for realizing fluoride deep treatment and recycling recovery comprises the following steps: S1, acquiring process data of a fluoride deep treatment device and a recycling recovery device in the operation process, continuously acquiring the process data, and outputting an original monitoring data sequence arranged in time sequence; S2, removing obvious noise in an original monitoring data sequence, carrying out segmentation processing on the monitoring data according to preset time windows, and comparing the segmentation processing result with preset stage division conditions to identify characteristic data for stage judgment in each time window; s3, repeatedly identifying corresponding characteristic data in a plurality of sections of adjacent time windows, and outputting corresponding stage identifiers when the characteristic data in the adjacent time windows meet corresponding stage division thresholds; s4, extracting feature information for semantic judgment from the process data based on the current stage identification, comparing the extracted feature information with preset semantic judgment conditions item by item, and outputting a corresponding semantic result according to the comparison result; s5, comparing the semantic result with a preset fluoride deep treatment target or a recycling recovery target, selecting or combining a control strategy from a strategy library, outputting a strategy parameter set comprising deep treatment strategy parameters or recycling recovery strategy parameters, and controlling and outputting a deep treatment process or recycling recovery process according to the strategy parameter set. In a preferred embodiment, in S1, an independent acquisition channel is set for process data, so that various process data enter a data acquisition flow in an independent acquisition channel mode, wherein the process data comprise six types of data, namely raw water fluoride ion concentration data, deep treatment effluent fluoride ion concentration data, suspended particle size distribution data, stirring motor current data, sedimentation liquid level height data and resource recovery mother liquor fluoride ion concentration data; Performing periodic sampling on independent acquisition channels corresponding to the six types of process data in a uniform sampling period, respectively acquiring the six types of process data in each sampling