Search

CN-122019902-A - Multistage characteristic value hierarchical rendering method for time sequence data of mass Internet of things

CN122019902ACN 122019902 ACN122019902 ACN 122019902ACN-122019902-A

Abstract

The invention discloses a multistage characteristic value layering rendering method of time sequence data of a mass Internet of things. The method comprises the steps of predefining multi-level time granularity at the back end, calculating the maximum value and the minimum value of original data in each time window at fixed time, storing the maximum value and the minimum value as a multi-layer characteristic value data set in a lasting mode, intelligently matching an optimal level and requesting characteristic values of the level according to the current view time span at the front end, and drawing a vertical line segment or an envelope line by the aid of the maximum value and the minimum value of each window at the front end to render. And when the user interacts, the data hierarchy is dynamically switched to realize smooth drill-down. According to the invention, through the core technology of pre-computing extremum, matching according to the need and envelope curve rendering, the front-end single rendering data point is strictly controlled within thousands of levels, meanwhile, the fluctuation range and key event characteristics of the data are completely reserved, the high-performance and high-fidelity Web visualization of billions-level time sequence data is realized, and the technical problem that the traditional scheme cannot achieve in performance, precision and interactive experience is solved.

Inventors

  • WANG ZHIFANG

Assignees

  • 苏州极奇信息科技有限公司

Dates

Publication Date
20260512
Application Date
20260129

Claims (8)

  1. 1. The multi-level characteristic value layering rendering method for the time sequence data of the mass Internet of things is characterized by comprising the following steps of: The method comprises the steps of S1, pre-calculating a process at the back end, namely, predefining N standard time granularity levels, wherein N is more than or equal to 2, each level corresponds to a fixed time interval length, for each level, carrying out sliding or fixed window segmentation on original index data in a time sequence database of the Internet of things at fixed time according to the corresponding time interval length, calculating at least one key characteristic value in each window, wherein the key characteristic value at least comprises the maximum value and the minimum value of original data in the window; S2, front end matching and request flow, wherein a front end visualization module automatically calculates the total time span delta T of the front end visualization module according to the current time range to be visualized set by a user, and selects an optimal target level from N standard time granularity levels according to a preset matching rule, wherein the matching rule enables the time interval length of the selected target level to be adapted to the total time span delta T so that the number of data points to be rendered is lower than a preset threshold value; And S3, a data supply and rendering process, wherein the front end initiates a data request to a background service, the request carries the identification of the target level and the time range to be visualized, the background service retrieves all data records falling in the time range from the corresponding characteristic value data set and returns the data records to the front end, and the front end performs graphic drawing by utilizing the corresponding maximum value and the corresponding minimum value of each time interval according to the received data records so as to restore the data trend in the time range to be visualized.
  2. 2. The multi-level eigenvalue hierarchical rendering method of time series data of the mass internet of things according to claim 1 is characterized in that the front end performs graphic drawing based on received data records, specifically, for each time interval, a vertical line segment is drawn in a graph coordinate system by taking time mark of the interval as an abscissa and taking a minimum value as a starting point and a maximum value as an end point, or connecting lines are performed according to the sequence of drawing all minimum value points and drawing all maximum value points in the interval first, so as to form an envelope line.
  3. 3. The multi-level eigenvalue hierarchical rendering method of time series data of the mass internet of things according to claim 1 or 2 is characterized by further comprising an interactive drill-down flow, wherein the interactive drill-down flow comprises the following steps of responding to drill-down interactive operation of a user on a visual interface, the drill-down interactive operation comprises frame selection operation or scaling operation, and the front-end matching and requesting flow and the data supplying and rendering flow of step S2 are re-executed according to a new time range determined by the drill-down interactive operation, wherein the total time span corresponding to the new time range is matched to a level with a thinner time interval length.
  4. 4. The multi-level eigenvalue hierarchical rendering method of time series data of mass internet of things according to claim 3, wherein when the drill-down interaction operation is a frame selection operation, if a time span of a frame selection area is smaller than a time interval length of a target level used for current rendering, drill-down is automatically triggered, and a level with a finer granularity of a next level is used as a new target level.
  5. 5. The multi-level eigenvalue hierarchical rendering method of time series data of the internet of things of claim 1, wherein N standard time granularity levels are in an exponential or multiple relationship, at least covering interval lengths in units of seconds, minutes, hours and days, and wherein the time interval length of the level of the finest granularity matches the acquisition frequency of the original data.
  6. 6. The multi-level eigenvalue hierarchical rendering method of time series data of mass internet of things according to claim 1, wherein in the back-end pre-calculation process of step S1, the calculated key eigenvalue further comprises at least one of average value, first value, last value and sample number.
  7. 7. The multi-level eigenvalue hierarchical rendering method of time series data of the mass internet of things according to claim 1 is characterized in that the preset threshold is 1000 data points, and the matching rule is that one level with the largest time interval length among all levels satisfying the quotient value obtained by dividing the total time span delta T by the preset threshold is selected as the target level.
  8. 8. The hierarchical multi-level eigenvalue rendering method of time series data of internet of things of claim 1, wherein the persistence storage creates an independent database table for each level, the database table at least comprises index ID, window start time, maximum value and minimum value fields, and the index ID and window start time are used as a joint main key.

Description

Multistage characteristic value hierarchical rendering method for time sequence data of mass Internet of things Technical Field The invention relates to the technical field of industrial Internet of things and Web front-end visualization, in particular to a multistage characteristic value hierarchical rendering method for time sequence data of a mass Internet of things. Background In an industrial internet of things system, a single device can generate tens to hundreds of pieces of time series data per second, and millions of data points are generated when a whole plant is operated for 24 hours. Conventional Web visualization schemes typically employ the following strategies: Full pull + front end downsampling: all the original data are transmitted to a browser, and sampling or aggregation is carried out by JavaScript; the defects are long network transmission time consumption, high memory occupation (easy breakdown) and slow first screen rendering. And (3) back-end real-time aggregation query: when the user zooms or drags each time, the GROUP BY time () aggregation query is launched to the database; The method has the defects of high concurrency, rapid increase of database pressure, high response delay and incapability of meeting smooth interaction. Fixed particle size prepolymerization: Only pre-calculating a single particle size (e.g., 1 minute mean); The defect that macroscopic trend and microscopic detail cannot be considered, and saw tooth distortion or information loss occurs when the drill is put down. In addition, existing schemes generally ignore the key role of "high and low points" on trend restoration—just averaging will smooth out the peaks/valleys, resulting in an alarm event that is not visible. The problem of data volume cannot be solved by directly transmitting the original extremum. Therefore, a layered feature value supply mechanism that combines precision, performance and interactive experience is needed. Disclosure of Invention The invention aims to solve the defects of the prior art and provides a multistage characteristic value layering rendering method for time sequence data of a mass Internet of things. In order to achieve the purpose, the technical scheme adopted by the invention is that the multi-level characteristic value layering rendering method of the time sequence data of the mass Internet of things comprises the following steps: The method comprises the steps of S1, pre-calculating a process at the back end, namely, predefining N standard time granularity levels, wherein N is more than or equal to 2, each level corresponds to a fixed time interval length, for each level, carrying out sliding or fixed window segmentation on original index data in a time sequence database of the Internet of things at fixed time according to the corresponding time interval length, calculating at least one key characteristic value in each window, wherein the key characteristic value at least comprises the maximum value and the minimum value of original data in the window; S2, front end matching and request flow, wherein a front end visualization module automatically calculates the total time span delta T of the front end visualization module according to the current time range to be visualized set by a user, and selects an optimal target level from N standard time granularity levels according to a preset matching rule, wherein the matching rule enables the time interval length of the selected target level to be adapted to the total time span delta T so that the number of data points to be rendered is lower than a preset threshold value; And S3, a data supply and rendering process, wherein the front end initiates a data request to a background service, the request carries the identification of the target level and the time range to be visualized, the background service retrieves all data records falling in the time range from the corresponding characteristic value data set and returns the data records to the front end, and the front end performs graphic drawing by utilizing the corresponding maximum value and the corresponding minimum value of each time interval according to the received data records so as to restore the data trend in the time range to be visualized. As a further description of the above technical solution: The front end performs graph drawing based on the received data record, specifically, for each time interval, a vertical line segment is drawn in a graph coordinate system by taking the time mark of the interval as an abscissa and taking the minimum value as a starting point and the maximum value as an end point, or connecting lines are performed according to the sequence of drawing all minimum value points in the interval and drawing all maximum value points to form an envelope line. As a further description of the above technical solution: The method further comprises an interactive drill-down flow, wherein the interactive drill-down flow comprises the steps of responding t