Full-performance detection method and system based on big data technology

Abstract

The invention discloses a full-performance detection method and system based on a big data technology, and relates to the technical field of equipment performance detection.The full-performance detection method comprises the steps that multi-dimensional performance data of equipment is collected, the multi-dimensional performance data is preprocessed, and a full-performance detection result is obtained based on the time-space characteristics of the preprocessed multi-dimensional performance data; the method comprises the following steps: carrying out dynamic weighted fusion to generate a unified input signal, constructing a space-time transfer model according to the input signal and historical state data of equipment, predicting the state of the equipment, calculating an abnormal score of the equipment by comparing the difference between the predicted state and real-time measurement data, and calculating the dynamic weight of the equipment based on the abnormal score and the multi-dimensional performance data. And generating a comprehensive performance score of the equipment, and matching the equipment maintenance priority suggestions. According to the invention, the comprehensiveness, the accuracy, the timeliness and the robustness of equipment performance detection are improved, the intelligence and the pertinence of equipment maintenance are realized, the equipment management level is effectively improved, the maintenance cost is reduced, and the service life of the equipment is prolonged.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of equipment performance detection, and particularly relates to a full performance detection method and system based on big data technology. BACKGROUND

[0002] With the wide application of industrial automation and intelligent equipment, equipment performance detection and state monitoring have become the key link to ensure stable operation and efficient production of equipment. In particular, in the power, manufacturing, communication, transportation and other industries, equipment failure or performance decline can cause serious economic losses or safety accidents, so it is crucial to accurately and comprehensively detect the performance of equipment.

[0003] However, the traditional equipment performance detection method has many shortcomings and cannot meet the requirements of modern intelligent equipment for detection accuracy, real-time performance and comprehensiveness. Existing detection methods usually rely on manual inspection, regular maintenance and traditional test equipment, which not only have low efficiency, but also are easily disturbed by human factors, making it difficult to achieve real-time monitoring and accurate prediction. In addition, the performance evaluation of traditional detection systems is based on single or limited indicators, which cannot fully reflect the overall health status of equipment and is difficult to cope with the dynamic changes of equipment in complex working environments.

[0004] The existing technology lacks dynamic adaptability in multi-dimensional data fusion, cannot effectively cope with complex temporal and spatial dynamic changes, and is prone to data distortion or inaccuracy. In terms of equipment state prediction, the model parameters cannot be dynamically adjusted in real time, making it difficult to adapt to the rapid changes of equipment operating state, which significantly limits the prediction accuracy, resulting in insufficient sensitivity of abnormal detection results, easy to miss or misreport equipment failure, lack of sufficient robustness and stability. The existing technology also has obvious shortcomings in the intelligentization and automation of detection process, which cannot meet the needs of modern equipment detection for efficiency, accuracy and comprehensiveness. SUMMARY

[0005] In view of the above existing problems, the present application provides a full performance detection method and system based on big data technology to solve the problems that the existing technology cannot effectively cope with complex temporal and spatial dynamic changes, is prone to data distortion or inaccuracy, and the prediction accuracy is significantly limited, resulting in insufficient sensitivity of abnormal detection results and easy to miss or misreport equipment failure.

[0006] To solve the above technical problems, a full performance detection method based on big data technology is proposed, which comprises,

[0007] The multi-dimensional performance data of the collection device is collected, the multi-dimensional performance data is preprocessed, a unified input signal is generated by dynamically weighting and fusing based on the spatio-temporal characteristics of the preprocessed multi-dimensional performance data, a time-space transfer model is constructed according to the input signal and historical state data of the device, the state of the device is predicted, the abnormal score of the device is calculated by comparing the difference between the predicted state and the real-time measurement data, the comprehensive performance score of the device is generated based on the abnormal score and the dynamic weight of the multi-dimensional performance data, and the device maintenance priority suggestion is matched.

[0008] As a preferred scheme of the full performance detection method based on big data technology, the multi-dimensional performance data includes electrical parameters, temperature, frequency and electromagnetic compatibility parameters.

[0009] The preprocessing of the multi-dimensional performance data includes sliding average filtering processing of the multi-dimensional performance data, identification and elimination of sudden abnormal data according to a preset threshold, and interpolation completion of missing data.

[0010] As a preferred scheme of the full performance detection method based on big data technology, the dynamic weighting and fusing includes dynamically weighting and fusing to generate a unified input signal based on the spatio-temporal characteristics of the preprocessed multi-dimensional performance data.

[0011] Generating a unified input signal includes dynamically calculating the weighting coefficient of each sensor data according to the time decay characteristics and spatial correlation of the sensor data, normalizing each sensor data according to the weighting coefficient, and generating an input signal.

[0012] As a preferred scheme of the full performance detection method based on big data technology, the construction of the time-space transfer model includes defining a state transition matrix according to the historical state data of the device, defining an input matrix combined with the influence of the input signal on the device state, and dynamically adjusting the parameters of the state transition matrix and the input matrix through a real-time feedback mechanism.

[0013] As a preferred scheme of the full performance detection method based on big data technology, the dynamic adjustment of the parameters of the state transition matrix and the input matrix includes calculating the adjustment amount of the state transition matrix and the input matrix based on the difference between the predicted state and the target state, and controlling the adjustment step through a preset learning rate to optimize the state transition matrix and the input matrix.

[0014] The target state is defined by the preset performance indicators and operation specifications of the device.

[0015] As a preferred scheme of the full performance detection method based on big data technology, the abnormal score of the computing device comprises: predicting the state of the device through a space-time transfer model, and calculating the abnormal score of the device by comparing the difference between the predicted state and the real-time measurement data.

[0016] The abnormal score of the computing device further comprises: performing weighted summation on the relative difference between the predicted value and the actual measured value of each performance indicator, introducing a standard deviation factor to smooth the weighted summation result, setting a dynamic threshold according to historical data statistical characteristics, and the formula is represented as:

[0017]

[0018] Wherein, G(t) is the abnormal score, i.e. the abnormal degree of the device at time t, s j (t) is the actual measured value of the jth performance indicator, is the value of the jth performance indicator obtained by prediction through the space-time transfer model, m is the number of performance indicators, is the weighted coefficient of the jth performance indicator, σ is the standard deviation, t is the time, j is the variable index, and when the abnormal score exceeds the predetermined threshold 1, the system triggers an alarm to prompt that the device has a failure and performance decline.

[0019] As a preferred scheme of the full performance detection method based on big data technology, the generation of the comprehensive performance score of the device comprises: generating the comprehensive performance score of the device based on the abnormal score and the dynamic weight of the multi-dimensional performance data, and matching the device maintenance priority suggestion.

[0020] The generation of the comprehensive performance score of the device further comprises: decomposing the performance indicators of the device, respectively calculating the dynamic weight of each sub-dimension, combining the interaction effect and time cumulative effect among the performance indicators, constructing a nonlinear scoring formula, and outputting the comprehensive performance score of the device through the scoring formula, and the formula is represented as:

[0021]

[0022] Wherein, Q(t) is the comprehensive performance score of the device at time t, G j (t) is the abnormal score of the jth performance indicator, β is the smoothing factor of the abnormal score, θ is the target threshold of the performance indicator, ρ jk is the interaction correlation coefficient between the performance indicators, i.e. the interaction effect weight of the jth and kth performance indicators, s k (t) is the actual measured value of the kth performance indicator, κ is the coefficient of the time penalty term, The sensitivity factor is dynamically adjusted for time, Δt is the time interval between the current time of the device and the last maintenance time, m is the number of performance indicators, j and k are variable indexes, is the weighting coefficient of the jth performance indicator, s j (t) is the actual measurement value of the jth performance indicator, t is the time.

[0023] As a preferred scheme of the full performance detection system based on big data technology, characterized in that, comprising a data acquisition and preprocessing module, a dynamic weighted fusion module, a construction of space-time transfer model and state prediction module and an abnormal detection and comprehensive evaluation module.

[0024] The data acquisition and preprocessing module is used for real-time acquisition of performance data of the device, and filtering and smoothing, abnormal value elimination and missing data completion of the original data.

[0025] The dynamic weighted fusion module is used for dynamically calculating the weighting coefficient of each sensor data based on the time decay factor and the spatial correlation factor, and generating a unified input signal through normalization processing and weighted summation.

[0026] The construction of space-time transfer model and state prediction module is used for constructing the device state transfer equation, initializing the state transfer matrix and the input matrix through the historical data, and dynamically adjusting the matrix parameters based on the difference between the predicted state and the target state.

[0027] The abnormal detection and comprehensive evaluation module is used for calculating the abnormal score, generating the comprehensive performance score, and triggering the hierarchical alarm and maintenance priority suggestion.

[0028] A computer device, comprising a memory and a processor, the memory stores a computer program, and the processor implements the steps of the method for full performance detection based on big data technology when executing the computer program.

[0029] A computer readable storage medium, which stores a computer program, and the computer program is executed by a processor to implement the steps of the method for full performance detection based on big data technology.

[0030] The present application has the beneficial effects that: the present application collects multi-dimensional performance data such as electrical parameters, temperature, frequency and electromagnetic compatibility parameters, and uses pre-processing means such as sliding average filtering, outlier rejection and missing data interpolation completion, effectively eliminates the interference of environmental noise and sensor errors on data, ensures that the input data of subsequent analysis has high reliability and continuity, and lays a foundation for accurate state evaluation; the weighting coefficients of sensor data are dynamically calculated based on time decay characteristics and spatial correlation, and a unified input signal is generated through normalization processing, solving the problem of unreasonable weight distribution caused by sensor position difference or insufficient data timeliness in traditional data fusion, significantly improving the precision and adaptability of multi-source data fusion, especially when the device state changes rapidly, the accuracy of data representation can still be maintained; by defining the state transition matrix and the input matrix, and dynamically adjusting the model parameters combined with the real-time feedback mechanism, the spatio-temporal dynamic modeling of the device state is realized, the model can adapt to the changes of the device operating environment, accurately predict the future state, and overcome the defect of large prediction deviation of the traditional static model under complex working conditions; by comparing the difference between the predicted state and the actual measured value, combining weighted summation and standard deviation smoothing processing to generate an abnormal score, and setting a dynamic threshold to trigger a graded alarm, the sensitivity of fault detection is improved, and false alarms caused by instantaneous fluctuations are reduced through nonlinear smoothing, realizing early and accurate identification of abnormal events; by decomposing the performance index sub-dimension, introducing interaction effect and time penalty term, a nonlinear comprehensive score formula is constructed to quantify the overall health status of the device, not only reflecting single index abnormalities, but also capturing the correlation between indexes and long-term performance decay trend, so as to generate a scientific and reasonable maintenance priority suggestion, optimizing the resource allocation efficiency. BRIEF DESCRIPTION OF DRAWINGS

[0031] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings needed in the embodiment description will be briefly introduced. Obviously, the drawings in the following description are only some embodiments of the present application, and other drawings can be obtained by those skilled in the art without creative labor.

[0032] Figure 1 The overall flowchart of a full performance detection method based on big data technology provided by an embodiment of the present application.

[0033] Figure 2 The system scheme flowchart of a full performance detection system based on big data technology provided by an embodiment of the present application. DETAILED DESCRIPTION

[0034] To make the above-mentioned objects, features, and advantages of the present invention more clearly understood, the following detailed description of the specific embodiments of the present invention is given in conjunction with the accompanying drawings. It is obvious that the described embodiments are only part of the embodiments of the present invention, not all of them. Based on the embodiments of the present invention, all other embodiments obtained by ordinary persons in this field without creative work should fall within the scope of protection of the present invention.

[0035] In the following description, many specific details are set forth to facilitate a full understanding of the present invention. However, the present invention may also be implemented in other ways different from those described herein. Those skilled in the art may make similar generalizations without violating the connotation of the present invention. Therefore, the present invention is not limited to the specific embodiments disclosed below.

[0036] Secondly, the term "one embodiment" or "embodiment" herein refers to a specific feature, structure, or characteristic that may be included in at least one implementation of the present invention. The phrase "in one embodiment" appearing in various places throughout this specification does not necessarily refer to the same embodiment, nor does it individually or selectively refer to an embodiment that is mutually exclusive of other embodiments.

[0037] The present invention is described in detail with reference to schematic diagrams. For ease of illustration, cross-sectional views of device structures may be partially enlarged and not to scale when describing embodiments of the present invention. Furthermore, the schematic diagrams are merely illustrative and should not limit the scope of the present invention. Furthermore, in actual production, the three-dimensional dimensions of length, width, and depth should be included.

[0038] In the description of the present invention, it should be noted that the terms "upper, lower, inner, and outer" and other references to orientations or positional relationships are based on the orientations or positional relationships shown in the accompanying drawings and are intended solely to facilitate and simplify the description of the present invention. They are not intended to indicate or imply that the devices or components referred to must have, be constructed, or operate in a specific orientation, and therefore should not be construed as limitations on the present invention. Furthermore, the terms "first, second, or third" are used for descriptive purposes only and should not be construed as indicating or implying relative importance.

[0039] In this disclosure, unless otherwise specified or limited, the terms "mounted," "connected," and "connected" should be interpreted broadly. For example, they may refer to fixed, removable, or integral connections. They may also refer to mechanical, electrical, or direct connections, indirect connections through an intermediary, or internal communication between two components. Those skilled in the art will understand the specific meanings of these terms in this disclosure.

[0040] Example 1, with reference to Figure 1 , which is the first embodiment of the present invention, provides a full performance detection method based on big data technology, including:

[0041] S1: Collect multi-dimensional performance data of the device, preprocess the multi-dimensional performance data, and generate a unified input signal based on the spatiotemporal characteristics of the preprocessed multi-dimensional performance data.

[0042] Further, multi-dimensional device performance data is collected, including but not limited to current, voltage, power, temperature, frequency, electromagnetic compatibility; in order to deal with noise, errors and lost data that occur during real-time collection, the original data collected by the sensor is preprocessed, including filtering to remove outliers and noise, effectively avoiding the influence of external environmental changes or sensor failure on the data.

[0043] Specifically, for voltage and current signals, a moving average filter is used to smooth the data, thereby reducing instability caused by short-term fluctuations; for the remaining types of signals, such as temperature or electromagnetic field strength, sudden fluctuations are determined and removed to ensure data continuity and reliability.

[0044] Further, after data collection, in order to ensure that the multi-dimensional data provided by different sensors can be effectively fused in time and space, the multi-dimensional device performance data needs to be weighted; each data point comes from different sensors, and the time characteristics and spatial location characteristics of the data need to be combined, so that the weighting coefficient of each data point at a specific time can be adjusted according to the current running state and historical data of the device.

[0045] Through weighting, the influence of each data can be adjusted when the device state changes, ensuring the accuracy of the data fusion result; the calculation method of the weighting coefficient considers a time decay factor and a spatial correlation factor, the time decay factor controls the degree of data decay over time, and the spatial correlation factor reflects the difference of data at different positions, and the spatiotemporal weighting coefficient of each data point is expressed as:

[0046]

[0047] where w i (t) is the weighting coefficient of the i-th sensor data at time t, λ i is the time decay factor of the i-th sensor data, which determines the speed of data decay over time, T i is the time of the last important event of the i-th sensor data, reflecting the real-time and historical influence of the data source, a i (t) is the spatial correlation factor of the i-th sensor data, which is adjusted according to the physical location of the device, reflecting the influence strength of sensors at different positions on the device state, n is the number of sensor data, i is the variable index, and t is the time.

[0048] Through the weighting process, different sensor data at the same time is reasonably weighted according to the space-time characteristics, and the data fused by weighting is calculated as the weighted sum of all sensor data, and the specific formula is as follows:

[0049]

[0050] Wherein, F(t) is the data fusion result at t moment, D i (t) is the measurement value of the i th sensor data at t moment, w i (t) is the weighting coefficient of the i th sensor data at t moment, n is the number of sensor data, i is the variable index, and t is the time.

[0051] In the embodiment of the application, the calculation of the weighting coefficient comprises dynamically calculating the weighting coefficient based on the time attenuation factor and the space correlation factor, and generating the input signal.

[0052] In an alternative embodiment, the calculation of the weighting coefficient comprises defining the sliding window length, calculating the weight based on the data in the window, dynamically adjusting the time attenuation factor according to the variance of the data in the window, calculating the space correlation factor through the topological relationship of the physical layout of the sensor, and generating the input signal.

[0053] In another alternative embodiment, the calculation of the weighting coefficient comprises clustering the historical multi-dimensional data, dividing the typical operating state of the device, predefining different weighting coefficient templates for each state, detecting the current state category of the device in real time, and calling the corresponding template to calculate the input signal.

[0054] S2: According to the input signal and the historical state data of the device, a space-time transfer model is constructed to predict the state of the device, and by comparing the difference between the predicted state and the real-time measurement data, the abnormal score of the device is calculated.

[0055] Further, in order to capture the dynamic characteristics of the device operating state in the time and space dimensions, and realize accurate state prediction and abnormal detection, a space-time transfer model of the device state is established to predict the future state of the device; The state of the device is composed of multiple performance indicators, such as voltage, current, temperature, and the change of the device in the current state not only depends on the historical state of the device, but also is affected by the fusion result of the sensor data, therefore, the state of the device can be described by the space-time transfer model, and the formula is as follows:

[0056] S(t+1)=A(t)·S(t)+B(t)·F(t)

[0057] Wherein, S(t+1) is the device state vector at t+1 time, S(t) is the device state vector at t time, A(t) is a state transition matrix, reflecting the transition rule of the current state of the device to the next time, which is obtained by historical data deduction, B(t) is an input matrix, describing the influence degree of external input data on the device state, F(t) is the data fusion result at t time, and t is the time.

[0058] In the embodiments of the present application, the prediction of the state of the device comprises predicting the device state through the state transition matrix and the input matrix, and dynamically adjusting the parameters.

[0059] In an alternative embodiment, the prediction of the state of the device comprises dividing the device life cycle into three stages of startup, steady state and decline, and automatically switching the stages according to the device running time and the comprehensive score.

[0060] In another alternative embodiment, the prediction of the state of the device comprises introducing attention weights in the adjustment of the state transition matrix and the input matrix, and calculating the similarity score of the historical state and the current input.

[0061] Furthermore, in order to improve the prediction ability of the space-time transition model, it is necessary to dynamically adjust the state transition matrix and the input matrix according to the gap between the current device state and the target state. In actual operation, the transition of the device state is not only affected by the previous state, but also adjusted according to the real-time running state of the device. Therefore, based on the error feedback mechanism, the state transition matrix and the input matrix are optimized in real time, and the formula is represented as:

[0062] A(t) = A(t-1) + τ·δ A (t)·(S target (t)-S(t))

[0063] B(t) = B(t-1) + τ·δ B (t)·(S target (t)-S(t))

[0064] Wherein, A(t) is a state transition matrix, B(t) is an input matrix, τ is a learning rate, controlling the pace of feedback adjustment, δ A (t) is the adjustment coefficient of the state transition matrix, δ B (t) is the adjustment coefficient of the input matrix, S target (t) is a target state vector, representing the performance index set of the device in the ideal state, provided by the device operation specification, and S(t) is the device state vector at t time.

[0065] It should be noted that the purpose of feedback adjustment is to ensure that the state transition process of the device can be optimized with the changes of real-time data, so that the spatio-temporal transition model prediction is always consistent with the actual state of the device, thereby the spatio-temporal transition model is effectively adjusted and optimized at each moment, improving the accuracy of prediction.

[0066] In order to identify possible failures or performance abnormalities of the device during operation, by comparing the difference between the actual state of the device and the model predicted state, it is judged whether the device is in an abnormal state; for this purpose, an abnormal score function is introduced, according to the difference between the actual measurement value of the device and the predicted value of the spatio-temporal transition model, combined with the weighted coefficient to evaluate the abnormality degree of the device at t time, the formula is expressed as:

[0067]

[0068] Wherein, G(t) is the abnormal score, i.e. the abnormality degree of the device at t time, s j (t) is the actual measurement value of the jth performance index, is the value of the jth performance index obtained by the spatio-temporal transition model prediction, m is the number of performance indexes, is the weighted coefficient of the jth performance index, σ is the standard deviation, t is the time, j is the variable index, when the abnormal score exceeds the predetermined threshold 1, the system triggers an alarm, indicating that the device has signs of failure and performance decline, when 1≤G(t)<1.5, a first level warning is triggered and a log is recorded, when G(t)≥1.5, a second level alarm is triggered and the maintenance personnel are notified.

[0069] In the embodiments of the present application, the abnormal score of the computing device includes the weighted sum of the relative difference between the predicted value and the actual value, combined with the standard deviation smoothing to generate the abnormal score.

[0070] In an alternative embodiment, the abnormal score of the computing device includes calculating the residual of each performance index, performing kernel density estimation on the historical residual, constructing a probability distribution model, and defining the abnormal score as the negative logarithmic probability of the current residual.

[0071] In another alternative embodiment, the abnormal score of the computing device includes defining a multi-time scale window, calculating the mean and variance of the prediction error in each window respectively, and fusing the multi-scale results to generate the abnormal score, and the dynamic threshold is adjusted according to the window statistical results.

[0072] S3: Based on the abnormal score and the dynamic weight of multi-dimensional performance data, a comprehensive performance score of the device is generated, and a device maintenance priority suggestion is matched.

[0073] Further, in order to quantify the current health status of the equipment, the overall health status of the equipment is evaluated according to the comprehensive score of the equipment, through multi-level index analysis and dynamic weight adjustment, a comprehensive evaluation formula is introduced, the evaluation of different performance indicators of the equipment is refined, and multiple sub-performance dimensions of the equipment are decomposed, and dynamic weighting and nonlinear effect are introduced in the comprehensive score, so that the detection requirements of the performance of the equipment are fully covered.

[0074] Further, the comprehensive score formula is represented as:

[0075]

[0076] Wherein, Q(t) is the comprehensive performance score of the equipment at time t, G j (t) is the abnormal score of the jth performance indicator, β is the smoothing factor of the abnormal score, θ is the target threshold of the performance indicator, ρ jk is the interaction correlation coefficient between performance indicators, that is, the interaction effect weight of the jth and kth performance indicators, s k (t) is the actual measurement value of the kth performance indicator, κ is the coefficient of the time penalty term, is the sensitivity factor of time dynamic adjustment, Δt is the time interval between the current time of the equipment and the last maintenance time, m is the number of performance indicators, j and k are variable indexes, is the weighting coefficient of the jth performance indicator, s j (t) is the actual measurement value of the jth performance indicator, t is the time.

[0077] The comprehensive performance score output is the performance comprehensive evaluation result of the current time of the equipment, and by comparing the comprehensive performance score values at different time points, the change trend of the performance of the equipment can be directly evaluated, and full performance detection is realized.

[0078] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present application and are not limiting, although the present application has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solutions of the present application can be modified or replaced equivalently without departing from the spirit and scope of the technical solutions of the present application, and they should be covered in the scope of the claims of the present application.

[0079] Embodiment 2, referring to Figure 2 is a second embodiment of the present application, which provides a full performance detection system based on big data technology, comprising a data acquisition and preprocessing module, a dynamic weighting fusion module, a construction of a space-time transfer model and a state prediction module, and an abnormality detection and comprehensive evaluation module.

[0080] The data acquisition and preprocessing module is used to acquire the performance data of the equipment in real time, and to filter and smooth the original data, remove outliers and complete missing data.

[0081] The dynamic weighting fusion module is configured to dynamically calculate the weighting coefficients of the sensor data based on a time decay factor and a spatial correlation factor, and generate a unified input signal through normalization processing and weighted summation.

[0082] The construction of the space-time transfer model and the state prediction module is configured to construct a device state transfer equation, initialize the state transfer matrix and the input matrix through historical data, and dynamically adjust the matrix parameters based on the difference between the predicted state and the target state.

[0083] The abnormality detection and comprehensive evaluation module is configured to calculate an abnormality score, generate a comprehensive performance score, and trigger a hierarchical alarm and maintenance priority suggestion.

[0084] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present application and are not limiting. Although the present application has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solutions of the present application can be modified or replaced equivalently without departing from the spirit and scope of the present application, and they should be covered in the scope of the claims of the present application.

[0085] Example 3 is a third embodiment of the present application, which provides a full performance detection method based on big data technology. In order to verify the beneficial effects of the present application, scientific demonstration is carried out through experiments.

[0086] In the smart grid system, the smart electric energy meter is the core equipment of power measurement, which can measure the power consumption of users in real time and upload data to the power grid dispatching center. However, in the long-term operation process, due to the influence of external environment, such as electromagnetic interference, temperature change, voltage fluctuation, even due to aging, loose line connection and other factors, the electric energy meter may have measurement deviation or even failure. Therefore, before the equipment is connected to the power supply company, the measurement performance of the smart electric energy meter needs to be fully detected to ensure that it meets the national measurement standard and can work stably under complex working conditions; the test equipment covers single-phase and three-phase smart electric energy meters, and the technical standards of the test environment are as follows:

[0087] Test voltage range: 190V-250V (actual rated value 220V), test current range: 0.1A-100A (different user load scenarios), power factor requirement: 0.8-1.0 (simulate different load characteristics), temperature range: -10℃-50℃ (adapt to different climate environments), electromagnetic compatibility standard: meet IEC 61000-4-2 specification, prevent external electromagnetic interference from affecting measurement accuracy, detection equipment: automatic detection platform, equipped with high-precision current, voltage, temperature and electromagnetic interference sensors, test object: 100 smart electric energy meters, detect their performance under different loads and environmental conditions

[0088] In this experiment, 100 intelligent electric energy meters are monitored synchronously, and the current, voltage, power, frequency, temperature and other data of each are collected. Through big data processing and intelligent analysis method, it is ensured that all electric energy meters meet the measurement accuracy requirements.

[0089] In data preprocessing, the collected signal is first smoothed to reduce the influence of high-frequency noise. The sliding average filter uses the following formula:

[0090]

[0091] The measurement values at the two moments are as follows:

[0092] Voltage historical data: [220.3V, 220.4V, 220.5V, 220.6V, 220.7V], filtered voltage value: (no abnormal fluctuation), similarly, smooth the current, power, temperature and other data to ensure data stability.

[0093] After data preprocessing, in order to better reflect the performance of electric energy meter under different working conditions, spatio-temporal weighted fusion is needed. Different sensor data may have different effects on the state of the device, so dynamic weighting is needed according to time decay and spatial correlation.

[0094] The spatio-temporal weighting coefficient of sensor data is calculated using the weighted formula:

[0095]

[0096] Where λ i = 0.05 is the time decay factor, T i = 5s is the last important event time, the spatial correlation factor α1(10) = 1.0, the normalized factor z(10) = 4.81 is calculated, and the following is obtained:

[0097]

[0098] The fusion data is obtained by weighting and summing all sensor data:

[0099]

[0100] The performance state of the device at the next moment is predicted using the state transition matrix and input matrix:

[0101] S(11) = A(10)·S(10) + B(10)·F(10)

[0102] The initial matrix is given:

[0103]

[0104] The calculation S(11) = [223.15, 15.48, 3206.78] is obtained.

[0105] The abnormal score is calculated by comparing the actual measurement value with the predicted value:

[0106]

[0107] Wherein, sigma = 0.5, phi j (11) = [0.3, 0.4, 0.3], the calculation G(11) = 0.92 is obtained, which is lower than the threshold 1.0, indicating that the equipment state is normal.

[0108] The comprehensive performance score of the intelligent electric energy meter is calculated:

[0109]

[0110] The calculation Q(11) = 976.8 is obtained, indicating that the detection equipment meets the measurement standard.

[0111] The obtained comprehensive score Q(11) = 976.8 is compared with the results of the existing detection method, and the experimental data analysis results are as follows:

[0112]

[0113] The detection method of the present application is superior to the traditional method in many aspects, wherein the abnormal detection accuracy is improved by 13.4%, the fault prediction ability reaches 92.3%, and the traditional method does not have fault prediction function at all; in addition, the accuracy of equipment qualification rate evaluation is improved by 8.4%, which shows that the present application can more accurately judge whether the equipment meets the measurement standard, and the detection time is shortened by 25 minutes, improving the test efficiency.

[0114] Embodiment 4, the fourth embodiment of the present application, which is different from the first three embodiments:

[0115] If the functions are implemented in the form of software function units and sold or used as independent products, they can be stored in a computer readable storage medium. Based on this understanding, the technical solutions of the present application or the part of the technical solutions that essentially contribute to the prior art or the part of the technical solutions can be embodied in the form of a software product. The computer software product is stored in a storage medium and includes a plurality of instructions for causing a computer device (which can be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the method described in the various embodiments of the present application. The aforementioned storage medium includes: a U disk, a mobile hard disk, a read-only memory (ROM, Read-Only Memory), a random access memory (RAM, Random Access Memory), a magnetic disk or an optical disk, and various media that can store program codes.

[0116] The logic and / or steps represented in the flowchart or otherwise described herein, for example, can be considered a list of executable instructions for implementing logic functions, and can be specifically embodied in any computer-readable medium for use by an instruction execution system, apparatus, or device, such as a computer-based system, a system including a processor, or other system that can fetch the instructions from the instruction execution system, apparatus, or device and execute the instructions, or in conjunction with these instructions execution systems, apparatuses, or devices. For the purpose of this specification, "computer-readable medium" can be any device that can contain, store, communicate, propagate, or transport programs for use by an instruction execution system, apparatus, or device, or in conjunction with these instruction execution systems, apparatuses, or devices.

[0117] More specific examples (a non-exhaustive list) of the computer-readable medium include the following: an electrical connection having one or more wires (electrical devices), a portable computer diskette (magnetic devices), a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber device, and a portable compact disc read-only memory (CDROM). In addition, the computer-readable medium can even be paper or other suitable medium on which the program can be printed, as the program can be electronically obtained, for example, by optical scanning of the paper or other medium, followed by editing, interpreting, or otherwise processing, if necessary, in other suitable ways to be electronically obtained, and then stored in the computer memory.

[0118] It should be understood that aspects of the application can be implemented in hardware, software, firmware or combinations thereof. In the embodiments described above, various steps or methods can be implemented, for example, by software or firmware stored in memory and executed by a suitable instruction execution system. For example, if implemented in hardware, and in another embodiment, any of the following technology, known in the art, or combinations thereof, can be used: discrete logic circuitry having logic gates for implementing logic functions upon an application of data signals, application specific integrated circuits having appropriate combinational logic gates, programmable gate arrays (PGA), field programmable gate arrays (FPGA), and the like.

Claims

1. A full performance detection method based on big data technology, characterized by: include, Collect multi-dimensional performance data of the equipment, pre-process the multi-dimensional performance data, and perform dynamic weighted fusion based on the spatiotemporal characteristics of the pre-processed multi-dimensional performance data to generate a unified input signal; Based on the input signal and the historical status data of the device, a spatiotemporal transfer model is constructed to predict the status of the device. By comparing the difference between the predicted status and the real-time measurement data, the abnormality score of the device is calculated; Based on the dynamic weighting of abnormality scores and multi-dimensional performance data, a comprehensive performance score for the equipment is generated and matched with equipment maintenance priority recommendations.

2. The full performance detection method based on big data technology according to claim 1, characterized in that: The multi-dimensional performance data includes electrical parameters, temperature, frequency and electromagnetic compatibility parameters; The preprocessing of the multi-dimensional performance data includes performing sliding average filtering on the multi-dimensional performance data, identifying and eliminating sudden abnormal data according to a preset threshold, and interpolating and completing missing data.

3. The full performance detection method based on big data technology according to claim 2, characterized in that: The dynamic weighted fusion includes generating a unified input signal by performing dynamic weighted fusion based on the spatiotemporal characteristics of the pre-processed multi-dimensional performance data; Generating a unified input signal includes dynamically calculating a weighting coefficient of each sensor data according to the time attenuation characteristics and spatial correlation of the sensor data, normalizing each sensor data according to the weighting coefficient, and generating an input signal.

4. The full performance detection method based on big data technology according to claim 3, characterized in that: The construction of the spatiotemporal transfer model includes defining a state transfer matrix based on historical state data of the device, defining an input matrix based on the impact of input signals on the device state, and dynamically adjusting parameters of the state transfer matrix and the input matrix through a real-time feedback mechanism.

5. The full performance detection method based on big data technology according to claim 4, characterized in that: Dynamically adjusting the parameters of the state transfer matrix and the input matrix includes calculating the adjustment amount of the state transfer matrix and the input matrix based on the difference between the predicted state and the target state, and controlling the adjustment pace through a preset learning rate to optimize the state transfer matrix and the input matrix; The target state is defined by the preset performance indicators and operating specifications of the equipment.

6. The full performance detection method based on big data technology according to claim 5, characterized in that: Calculating the anomaly score of the device includes predicting the state of the device using a spatiotemporal transfer model, and calculating the anomaly score of the device by comparing the difference between the predicted state and real-time measurement data; The abnormality scoring of the computing device also includes weighted summing of the relative differences between the predicted value and the actual measured value of each performance indicator, introducing a standard deviation factor to smooth the weighted summation result, and setting a dynamic threshold based on the statistical characteristics of historical data. The formula is expressed as follows: Among them, G(t) is the abnormality score, that is, the abnormality degree of the device at time t, s j (t) is the actual measured value of the jth performance indicator, is the value of the jth performance indicator predicted by the spatiotemporal transfer model, m is the number of performance indicators, is the weighting coefficient of the jth performance indicator, σ is the standard deviation, t is the time, and j is the variable index. When the anomaly score exceeds the predetermined threshold, the system triggers an alarm, indicating that the equipment has signs of failure and performance degradation.

7. The full performance detection method based on big data technology according to claim 6, characterized in that: Generating a comprehensive performance score for the equipment includes generating a comprehensive performance score for the equipment based on the anomaly score and dynamic weights of multi-dimensional performance data, and matching equipment maintenance priority recommendations; Generating a comprehensive performance score for a device further includes decomposing the performance indicators of the device, calculating the dynamic weight of each sub-dimension, combining the interaction effect between the performance indicators and the time accumulation effect, constructing a nonlinear scoring formula, and outputting the comprehensive performance score of the device through the scoring formula, which is expressed as follows: Among them, Q(t) is the comprehensive performance score of the equipment at time t, G j (t) is the anomaly score of the jth performance indicator, β is the smoothing factor of the anomaly score, θ is the target threshold of the performance indicator, and ρ jk is the interaction correlation coefficient between performance indicators, i.e., the interaction effect weight of the jth and kth performance indicators, s k (t) is the actual measured value of the kth performance indicator, κ is the coefficient of the time penalty term, is the sensitivity factor of dynamic time adjustment, Δt is the time interval between the current time and the last maintenance time of the equipment, m is the number of performance indicators, j and k are variable indexes, is the weighted coefficient of the jth performance indicator, s j (t) is the actual measured value of the jth performance indicator, and t is the time.

8. A system using the full performance detection method based on big data technology as claimed in any one of claims 1 to 7, characterized in that: It includes data acquisition and preprocessing module, dynamic weighted fusion module, spatiotemporal transfer model building and state prediction module, and anomaly detection and comprehensive evaluation module; The data acquisition and preprocessing module is used to collect the performance data of the equipment in real time, and filter and smooth the raw data, eliminate outliers and complete missing data; The dynamic weighted fusion module is used to dynamically calculate the weight coefficients of each sensor data based on the time attenuation factor and the spatial correlation factor, and generate a unified input signal through normalization processing and weighted summation; The spatiotemporal transfer model and state prediction module are constructed to construct the device state transfer equation, initialize the state transfer matrix and input matrix through historical data, and dynamically adjust the matrix parameters based on the difference between the predicted state and the target state; The anomaly detection and comprehensive evaluation module is used to calculate anomaly scores, generate comprehensive performance scores, and trigger graded alarms and maintenance priority recommendations.

9. A computer device comprising a memory and a processor, wherein the memory stores a computer program, wherein: When the processor executes the computer program, the processor implements the steps of a full performance detection method based on big data technology as described in any one of claims 1 to 7.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of a full performance detection method based on big data technology according to any one of claims 1 to 7 are implemented.

Images