Processing method and system for industrial asset positioning data collection based on multi-mode positioning

Abstract

The application relates to the field of communication positioning technology and discloses a processing method and system for industrial asset positioning data collection based on multi-mode positioning, which comprises the following steps: acquiring positioning data clusters, setting independent matching time windows for each positioning data cluster, and structurally integrating discrete data; comparing the positioning data in the matching time windows in terms of time, collecting and forming offset trend data, and calculating standard offset values; setting the calibrated standard offset values as reference standards, calculating other offset values, identifying positioning data with offset values exceeding a spatial tolerance threshold as abnormal positioning data, and removing the abnormal positioning data. The application avoids logical confusion and interference caused by mixed processing of data in different working states, provides a data basis for identifying and removing abnormal positioning data, and overcomes the defects of low analysis efficiency and insufficient accuracy caused by lack of effective preprocessing.

Description

Technical Field

[0001] This invention belongs to the field of communication and positioning technology, specifically relating to a processing method and system for industrial asset positioning data acquisition based on multi-mode positioning. Background Technology

[0002] With the development of big data and IoT technologies, industrial asset management technology based on multimodal positioning has improved the management efficiency and production flexibility of industrial assets. It can respond and handle risks to industrial assets in the first instance. Since industrial assets are mostly large equipment with relatively low update frequency, accurate positioning and movement trajectory collection of industrial assets can effectively prevent the loss of industrial assets.

[0003] Currently, industrial asset positioning technology based on multimodal positioning is easily affected by sensor errors and environmental interference during the data acquisition process. At the same time, due to the inherent defects of multimodal data fusion, it is easy to generate data offset problems with systematic or dynamic characteristics, resulting in deviations between the positioning data and the actual location. It is difficult to accurately judge and calibrate the data offset through conventional calibration or correction methods.

[0004] Current conventional threshold judgment or filtering methods are difficult to effectively distinguish between real data fluctuations and offsets. This makes it easy to misjudge valid data as abnormal and erroneously remove it during data cleaning and preprocessing. It is also easy to fail to identify and remove abnormal offset data, which affects the reliability and accuracy of positioning results. This can easily cause errors in the real-time location of industrial assets, misleading production decisions and delaying fault diagnosis, and adversely affecting the safe and efficient operation of industrial sites.

[0005] Therefore, this application provides a processing method and system for industrial asset positioning data acquisition based on multi-mode positioning. Summary of the Invention

[0006] The purpose of this invention is to provide a method and system for processing industrial asset positioning data acquisition based on multi-mode positioning.

[0007] To achieve the above-mentioned objectives, the technical solution adopted by the present invention is as follows:

[0008] The processing method for industrial asset location data acquisition based on multi-mode positioning includes the following steps:

[0009] When it is necessary to derive trend parameters for determining whether there are risky working conditions in a monitored area based on a set of positioning data clusters containing positioning data based on visual images and positioning data based on wireless signals, the following steps are performed: calibrate the standard offset value; identify and remove abnormal positioning data in the positioning data cluster based on the standard offset value to obtain valid positioning data; and derive the trend parameters using the valid positioning data.

[0010] The calibration standard offset value includes: setting a matching time window based on the time parameters in the positioning data cluster; within the matching time window, comparing the positioning data over time to form offset trend data; and calculating the standard offset value based on the offset trend data.

[0011] The step of identifying and removing abnormal positioning data from the positioning data cluster based on the standard offset value to obtain valid positioning data includes: setting the standard offset value as the reference standard; calculating the offset between other offset values ​​and the reference standard; and when the offset exceeds the spatial tolerance threshold, identifying the corresponding positioning data as abnormal positioning data and removing it.

[0012] Preferably, setting the matching time window based on the time parameters in the positioning data cluster includes:

[0013] Before the start time of the location data cluster, a first time node is selected as the upper limit time node, and after the start time node, a second time node is selected as the lower limit time node, so that the time period between the upper limit time node and the lower limit time node is set as the matching time window.

[0014] Preferably, setting the matching time window based on the time parameters in the positioning data cluster further includes:

[0015] When multiple time periods are available to set a matching time window, the time period with the smallest time difference from the current time period is selected as the priority window;

[0016] If the time parameters of the location data match the median of the priority window, the location data is classified into the priority window and marked as primary data; if the time parameters of the location data do not match the median of the priority window, it is classified into other time periods that serve as backup windows and marked as secondary data.

[0017] Preferably, the step of summarizing offset trend data by comparing the positioning data over time within the matching time window includes:

[0018] The time sequence of the master data is used as a benchmark; the location data is arranged one by one according to the benchmark for time comparison; the response interval of the delayed data is statistically analyzed as the delay offset value, and the offset trend data is summarized.

[0019] Preferably, the calculation of the standard offset value based on the offset trend data includes:

[0020] Based on the positive and negative distribution trend of the offset values ​​within each matching time window, the offset value with the highest frequency is calibrated as the standard offset value.

[0021] The industrial asset location data acquisition and processing system based on multi-mode positioning uses the aforementioned multi-mode positioning-based industrial asset location data acquisition and processing method, and includes the following modules:

[0022] The data acquisition and classification module is used to acquire location data generated by multiple mobile devices under different working states, including location data based on visual images and location data based on wireless signals, and to classify and summarize the location data to form location data clusters.

[0023] The abnormal data identification module is used to calibrate the standard offset value; based on the standard offset value, it identifies and removes abnormal positioning data in the positioning data cluster to obtain valid positioning data.

[0024] The trend parameter export module is used to export trend parameters using valid location data.

[0025] The risk condition assessment module is used to analyze trend parameters to determine whether there are risk conditions in the monitored area.

[0026] Preferably, the calibration standard offset value includes:

[0027] Based on the time parameters in the positioning data cluster, a matching time window is set; within the matching time window, the positioning data is compared over time to form offset trend data; based on the offset trend data, the standard offset value is calculated.

[0028] Preferably, the step of identifying and removing abnormal positioning data from the positioning data cluster based on standard offset values ​​to obtain valid positioning data includes:

[0029] Set the standard offset value as the baseline standard; calculate the offset between other offset values ​​and the baseline standard; when the offset exceeds the spatial tolerance threshold, identify the corresponding positioning data as abnormal positioning data and remove it.

[0030] Preferably, setting the matching time window based on the time parameters in the positioning data cluster includes:

[0031] Before the start time of the location data cluster, a first time node is selected as the upper limit time node, and after the start time node, a second time node is selected as the lower limit time node, so that the time period between the upper limit time node and the lower limit time node is set as the matching time window.

[0032] Preferably, setting the matching time window based on the time parameters in the positioning data cluster further includes:

[0033] When multiple time periods are available to set a matching time window, the time period with the smallest time difference from the current time period is selected as the priority window;

[0034] If the time parameters of the location data match the median of the priority window, the location data is classified into the priority window and marked as primary data; if the time parameters of the location data do not match the median of the priority window, it is classified into other time periods that serve as backup windows and marked as secondary data.

[0035] Beneficial effects

[0036] This invention acquires and categorizes location data, including information based on visual images and wireless signals, to form location data clusters. It sets independent matching time windows for each location data cluster, and integrates discrete data in a structured manner. This avoids logical confusion and interference that may be caused by mixing data from different working states, and provides a data foundation for identifying and eliminating abnormal location data. It overcomes the shortcomings of low analysis efficiency and insufficient accuracy caused by the lack of effective preprocessing.

[0037] This invention performs time comparison on location data within a matching time window, summarizes the offset trend data, and calculates a standard offset value that reflects the systematicness and regularity within a specific time period. It can distinguish between occasional abnormal location data and consistent delayed response data, avoids the accidental deletion of valuable data, and improves data utilization.

[0038] This invention sets the calibrated standard offset value as the benchmark standard and calculates other offsets. Positioning data with offsets exceeding the spatial tolerance threshold are identified as abnormal positioning data and removed. A dynamic and adaptive abnormal data evaluation standard based on the inherent laws of the data itself is established. The evaluation standard is adjusted according to the actual data performance under different working conditions, so as to achieve more effective identification and filtering of abnormal positioning data and ensure the accuracy of the trend parameters output by trend analysis. Attached Figure Description

[0039] Figure 1 This is a flowchart of the method provided by the present invention;

[0040] Figure 2 This is a system module diagram provided by the present invention. Detailed Implementation

[0041] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0042] Example 1

[0043] like Figure 1 As shown, this embodiment provides a processing method for industrial asset positioning data acquisition based on multi-mode positioning, including the following steps:

[0044] S1. Acquire and preprocess positioning data; collect positioning data generated by multiple mobile devices deployed in the industrial site under different working conditions, including positioning data based on visual images and positioning data based on wireless signals;

[0045] Specifically, visual image-based positioning data can be obtained by processing on-site video streams collected by image sensors such as industrial cameras, and determining the position information of the target asset in a preset coordinate system through image processing such as target recognition and attitude calculation; wireless signal-based positioning data can be obtained by receiving and parsing signals generated by wireless communication technologies such as UWB, RFID, or Bluetooth beacons, and calculating the position of the target asset as the signal source based on principles such as time difference of arrival, angle of arrival, or received signal strength indication.

[0046] After the data collection is completed, all the obtained positioning data points are globally sorted according to their respective timestamps, and the time intervals between adjacent data points are checked.

[0047] When the time interval between any two consecutive data points exceeds the preset activity interval threshold, it is determined that the continuous movement trajectory of the asset is interrupted at this point, and this time point is taken as the cutoff point; the preset activity interval threshold can be set to 5 minutes according to the specific industrial scenario.

[0048] By dividing all the location data using all identified cutoff points, the original continuous data stream is segmented and classified into several independent location data clusters. Each location data cluster corresponds to the movement trajectory of the asset within a certain continuous working period, thereby isolating irrelevant data interference for targeted analysis.

[0049] S2. Set a matching time window for each location data cluster. The matching time window is used to obtain the effective time range. The effective time range is used for correlation analysis. The earliest timestamp among all data points contained in a certain location data cluster is used as the start time point, and the latest timestamp is used as the end time point.

[0050] The upper limit time point is obtained by shifting the start time point forward by a preset first time amount, and the lower limit time point is obtained by shifting the end time point backward by a preset second time amount; wherein, both the preset first time amount and the preset second time amount can be set to 2 minutes;

[0051] The closed time period consisting of the upper limit time point and the lower limit time point is defined as the matching time window of the location data cluster; when the activity records of industrial assets present multiple discontinuous time periods, multiple location data clusters are formed.

[0052] Prioritize processing the location data cluster that is closest to the current time and its corresponding matching time window, and use it as the priority window. Data points within the location data cluster corresponding to the priority window are marked as primary data, and other data points within the priority window but not belonging to the corresponding location data cluster are marked as secondary data. The secondary data comes from other non-continuous activity segments and can be used as supplementary information for analysis.

[0053] S3. Perform time continuity verification on the data within the matching time window. Within the positioning data cluster composed of the master data, compare the response interval between adjacent data points with the preset response time threshold in chronological order. The preset response time threshold can be set to 200 milliseconds.

[0054] If the response interval is less than the preset response time threshold, it indicates that the time continuity between data points is normal, and the data point is marked as normal response data. If the response interval exceeds the response time threshold, it indicates that there is a delay or interruption in data acquisition, and the next data point among the adjacent data points is marked as delayed response data. For each data point marked as delayed response, the time difference between it and the previous data point is calculated as the delay offset value, which quantifies the degree of delay in data acquisition.

[0055] This results in the output of location information datasets with time continuity markers, which are then used as offset trend data to provide data for spatial location accuracy analysis.

[0056] S4. Perform abnormal data identification, calculate the standard offset value, traverse all positioning data within the matching time window, and calculate the offset of each data point relative to its previous valid data point.

[0057] The offset calculation process includes: based on the position, velocity and timestamp of the previous valid data point, the predicted position at the timestamp of the current data point is calculated, then the Euclidean distance between the actual recorded position of the current data point and the predicted position is calculated, and the Euclidean distance is used as the offset of the data point. The offset is used to quantify the degree of deviation of a single positioning result.

[0058] After collecting the offsets of all data points within the window, the offset values ​​are statistically processed. A standard offset value that can represent the inherent or environmentally caused general offset characteristics of the positioning within this period can be determined by constructing a histogram of the numerical distribution and identifying the interval where the data is most concentrated, or by calculating the median or mode of all offsets.

[0059] The set of positioning data points whose offset is close to the standard offset value is established as the reference value set, which represents the normal positioning behavior in this environment.

[0060] S5. Based on the standard offset value, perform anomaly judgment and filtering on the positioning data in the window; check the offset of other positioning data in the window one by one, and for the positioning data to be checked, calculate the absolute difference between its offset and the standard offset value used as the reference standard.

[0061] If the absolute difference does not exceed the preset spatial tolerance threshold, the positioning data is determined to be valid positioning data, indicating that its offset is within an acceptable error range.

[0062] If the absolute difference exceeds the spatial tolerance threshold, the positioning data is determined to be abnormal positioning data and is removed from subsequent trajectory analysis or specially marked.

[0063] The space tolerance threshold can be dynamically adjusted according to the physical size of the asset, the movement distance per second, and the accuracy requirements. It can be fixed at 3 meters or set as a specific multiple of the standard offset value. The space tolerance threshold can be calculated by weighting the physical size of the asset, the movement distance per second, and the accuracy requirements. The weights can be adjusted according to the accuracy requirements. The specific weights can be 0.3, 1, and 1.5 to calculate the space tolerance threshold.

[0064] Similar anomaly detection and screening are performed on the auxiliary data, but a more lenient auxiliary data location threshold is usually adopted. The auxiliary data location threshold is greater than the spatial tolerance threshold and can be fixed at 5 meters. If a set of continuous data points is found in the auxiliary data whose offset deviates significantly from the baseline standard of the main data, this set of auxiliary data is promoted to a new independent location data cluster, and a new matching time window is generated for it, starting a new round of analysis process, thereby achieving adaptive discovery of newly emerging unexpected asset activity patterns.

[0065] S6. Summarize all the selected valid positioning data and output the trend parameters. Reorganize all the valid positioning data that have been determined to be valid in chronological order to construct a highly reliable industrial asset movement trajectory that has been purified and corrected.

[0066] Based on the movement trajectory of industrial assets, a series of trend parameters with clear physical meaning are derived by calculating the positional changes and time intervals between adjacent valid data points. These trend parameters may include the instantaneous speed, average speed, direction of movement, angular velocity, acceleration, etc. of the industrial assets, as well as the dwell time or frequency of occurrence in a specific area.

[0067] These trend parameters can be directly output to upper-level management to support higher-level applications; specifically, they can be applied to production cycle analysis and logistics scheduling optimization, and can also serve as a basis for judging whether equipment is experiencing abnormal operation or entering dangerous areas, thereby triggering corresponding safety warning mechanisms.

[0068] Example 2

[0069] like Figure 2 As shown, this embodiment provides a processing system for industrial asset positioning data acquisition based on multi-mode positioning, including the following modules:

[0070] The data acquisition and classification module is used to access various positioning information sources in the industrial field, such as industrial cameras, UWB base stations, RFID readers, etc., to acquire raw positioning data based on visual images and positioning data based on wireless signals. Based on the continuity of data timestamps, it performs cutoff point identification and data segmentation to generate structured positioning data clusters representing different continuous activity cycles.

[0071] The abnormal data identification module is used to receive the generated location data clusters, and define the matching time window for correlation analysis for each location data cluster by applying a preset time offset based on the start and end time of each location data cluster. When there are multiple data clusters, the module determines the primary data to be processed and the supplementary auxiliary data to be processed based on the timeliness principle.

[0072] Within the matching time window, the time continuity of the master data is checked, the delayed response data is identified and its delay offset value is recorded, and the offset of all data points relative to the predicted position is calculated.

[0073] Through pre-defined statistical processing, such as calculating the mode or median of the offset, a standard offset value that can reflect the positioning deviation in the current environment is determined, and a corresponding set of benchmark reference values ​​is established.

[0074] Based on the standard offset value and the preset spatial tolerance threshold, each positioning data in the window is compared and judged one by one, and abnormal positioning data with abnormal offset is identified and removed. At the same time, it has the ability to discover new activity patterns in the auxiliary data, thereby outputting valid positioning data.

[0075] The trend parameter export module is used to integrate all verified valid positioning data, reconstruct the movement trajectory of industrial assets in time sequence, and calculate and output a series of quantitative trend parameters such as speed, direction, and acceleration based on the movement trajectory of industrial assets for production management or safety monitoring.

[0076] The risk condition judgment module is used to continuously receive trend parameters and compare them in real time with preset safety rules such as maximum speed limit and prohibited areas to determine whether there are risk conditions in the monitored area.

[0077] Once an asset's operational status is found to deviate from normal or violate rules, an alarm mechanism is triggered, sending a warning message to the management personnel.

[0078] Obviously, those skilled in the art can make various modifications and variations to this invention without departing from it. Therefore, if these modifications and variations fall within the scope of the claims of this invention and their equivalents, this invention also intends to include these modifications and variations.

Claims

1. A method for processing industrial asset location data acquisition based on multi-mode positioning, characterized in that, Includes the following steps: When it is necessary to derive trend parameters for determining whether there are risky working conditions in a monitored area based on a set of positioning data clusters containing positioning data based on visual images and positioning data based on wireless signals, execute: Calibrate standard offset value. Abnormal positioning data in the positioning data cluster is identified and removed based on standard offset values ​​to obtain valid positioning data; trend parameters are then derived using the valid positioning data. The calibration standard offset value includes: setting a matching time window based on the time parameters in the positioning data cluster; within the matching time window, comparing the positioning data over time to form offset trend data; and calculating the standard offset value based on the offset trend data. The step of identifying and removing abnormal positioning data from the positioning data cluster based on the standard offset value to obtain valid positioning data includes: setting the standard offset value as the reference standard; calculating the offset between other offset values ​​and the reference standard; and when the offset exceeds the spatial tolerance threshold, identifying the corresponding positioning data as abnormal positioning data and removing it.

2. The processing method for industrial asset positioning data acquisition based on multi-mode positioning according to claim 1, characterized in that, The step of setting a matching time window based on the time parameters in the positioning data cluster includes: Before the start time of the location data cluster, a first time node is selected as the upper limit time node, and after the start time node, a second time node is selected as the lower limit time node, so that the time period between the upper limit time node and the lower limit time node is set as the matching time window.

3. The processing method for industrial asset positioning data acquisition based on multi-mode positioning according to claim 2, characterized in that, The step of setting the matching time window based on the time parameters in the positioning data cluster further includes: When multiple time periods are available to set a matching time window, the time period with the smallest time difference from the current time period is selected as the priority window; If the time parameters of the location data match the median of the priority window, the location data is classified into the priority window and marked as primary data; if the time parameters of the location data do not match the median of the priority window, it is classified into other time periods that serve as backup windows and marked as secondary data.

4. The processing method for industrial asset positioning data acquisition based on multi-mode positioning according to claim 1, characterized in that, Within the matching time window, the process of comparing positioning data over time to generate offset trend data includes: The time sequence of the master data is used as a benchmark; the location data is arranged one by one according to the benchmark for time comparison; the response interval of the delayed data is statistically analyzed as the delay offset value, and the offset trend data is summarized.

5. The processing method for industrial asset positioning data acquisition based on multi-mode positioning according to claim 1, characterized in that, The calculation of the standard offset value based on the offset trend data includes: Based on the positive and negative distribution trend of the offset values ​​within each matching time window, the offset value with the highest frequency is calibrated as the standard offset value.

6. A processing system for industrial asset location data acquisition based on multi-mode positioning, comprising the processing method for industrial asset location data acquisition based on multi-mode positioning as described in any one of claims 1-5, characterized in that, Includes the following modules: The data acquisition and classification module is used to acquire location data generated by multiple mobile devices under different working states, including location data based on visual images and location data based on wireless signals, and to classify and summarize the location data to form location data clusters. Anomaly data identification module is used to calibrate standard offset values; Abnormal positioning data in the positioning data cluster is identified and removed based on standard offset values ​​in order to obtain valid positioning data. The trend parameter export module is used to export trend parameters using valid location data. The risk condition assessment module is used to analyze trend parameters to determine whether there are risk conditions in the monitored area.

7. The processing system for industrial asset positioning data acquisition based on multi-mode positioning according to claim 6, characterized in that, The calibration standard offset value includes: Based on the time parameters in the positioning data cluster, a matching time window is set; within the matching time window, the positioning data is compared over time to form offset trend data; based on the offset trend data, the standard offset value is calculated.

8. The processing system for industrial asset positioning data acquisition based on multi-mode positioning according to claim 6, characterized in that, The step of identifying and removing abnormal positioning data from positioning data clusters based on standard offset values ​​to obtain valid positioning data includes: Set the standard offset value as the baseline standard; calculate the offset between other offset values ​​and the baseline standard; when the offset exceeds the spatial tolerance threshold, identify the corresponding positioning data as abnormal positioning data and remove it.

9. The processing system for industrial asset positioning data acquisition based on multi-mode positioning according to claim 6, characterized in that, The step of setting a matching time window based on the time parameters in the positioning data cluster includes: Before the start time of the location data cluster, a first time node is selected as the upper limit time node, and after the start time node, a second time node is selected as the lower limit time node, so that the time period between the upper limit time node and the lower limit time node is set as the matching time window.

10. The processing system for industrial asset positioning data acquisition based on multi-mode positioning according to claim 9, characterized in that, The step of setting the matching time window based on the time parameters in the positioning data cluster further includes: When multiple time periods are available to set a matching time window, the time period with the smallest time difference from the current time period is selected as the priority window; If the time parameters of the location data match the median of the priority window, the location data is classified into the priority window and marked as primary data; if the time parameters of the location data do not match the median of the priority window, it is classified into other time periods that serve as backup windows and marked as secondary data.

Images