Real-time monitoring method for communication performance of electric energy terminal based on multi-dimensional pressure test

Abstract

The application discloses a kind of based on multi-dimensional pressure test electric energy terminal communication performance real-time monitoring method, including real-time acquisition terminal downlink communication, uplink communication and the multi-dimensional data of ontology operation;Characteristic extraction and index calculation are carried out to the collected data, cover physical layer signal quality, link layer communication efficiency, network layer connectivity and terminal resource state;Comprehensive performance index model is constructed, weight is distributed using analytic hierarchy process and is normalized, and quantitative index representing terminal communication state is calculated;According to the monitoring result or operation and maintenance demand triggers hierarchical pressure test, simulates normal, high load and recovery stage operating condition, dynamically evaluates the performance of terminal under extreme stress.This application solves the problem that the prior art lacks real-time online monitoring and performance evaluation mechanism under pressure state, realizes all-round quantitative evaluation and early warning to terminal communication performance, effectively improves the communication guarantee capability and operation and maintenance efficiency under the background of electric power spot trading.

Description

Technical Field

[0001] This invention relates to the field of power system communication monitoring technology. Specifically, it relates to a method for real-time monitoring of the communication performance of power energy terminals based on multi-dimensional stress testing. This method is particularly suitable for comprehensive and real-time evaluation and early warning of the communication quality of power energy acquisition terminals in actual operating environments under the background of power spot trading. Background Technology

[0002] With the continued deepening of power market reforms, electricity spot trading has placed higher demands on the real-time performance, accuracy, and reliability of electricity metering data. As a crucial intermediate node connecting smart meters and the master station system, the stability of the communication performance of the electricity data acquisition terminal directly affects the fairness of transaction settlement and the reliability of grid operation. In electricity spot trading scenarios, transaction clearing and settlement processes place stringent requirements on data transmission latency, integrity, and concurrent processing capabilities. Any communication interruption or latency jitter can lead to data loss, thereby affecting the accuracy of transaction results.

[0003] Currently, the power industry has established a relatively complete grid access testing mechanism for power energy collection terminals. Standardized testing procedures are used to quantitatively evaluate the communication performance of terminals under various operating conditions, providing a reference for equipment selection and bidding. However, the existing technical system still has significant limitations and shortcomings in actual operation monitoring:

[0004] First, existing technologies lack effective online monitoring methods for the real-time communication performance of terminals. Current monitoring methods mainly focus on the calibration of metering accuracy and the passive diagnosis of hardware faults. For example, some proposed methods for diagnosing communication faults in electricity meters only locate communication faults in individual meters and cannot assess the terminal's performance as a convergence node in terms of overall communication processing capabilities. Other proposed concentrator communication module testing devices are typically offline tools, requiring disassembly from the field or connection to specialized testing fixtures, making real-time monitoring under operating conditions impossible. Furthermore, online monitoring methods for electricity meters based on state estimation focus on analyzing the error characteristics of metering data, without addressing key performance parameters such as the physical signal quality and transmission delay of the communication link. In actual operation, communication link performance degradation (such as increased response latency, higher packet loss rate, and decreased concurrent processing capacity) often occurs gradually, usually only being passively discovered after data loss or settlement anomalies occur. The lack of early warning mechanisms for communication quality makes it difficult for maintenance personnel to grasp the communication status of terminals in the actual operating environment in real time, resulting in delayed problem response and difficulty in tracing fault origins.

[0005] Secondly, existing technologies lack a mechanism for evaluating terminal performance under stress. Electricity spot trading often experiences sudden data reporting peaks at critical settlement times such as the hour and half-hour, requiring terminals to handle a large number of concurrent requests within a short period, resulting in significant communication pressure. However, existing online monitoring methods do not consider the impact of load fluctuations on terminal performance and cannot assess the terminal's processing capacity and stability under extreme conditions. Even some research involving communication quality monitoring of power data acquisition systems often only monitors single-dimensional indicators (such as online rate or signal strength), lacking a systematic, multi-dimensional performance evaluation system. Terminals from the same manufacturer and of the same model often exhibit significant differences in performance under stress due to complex environmental factors such as installation location, line quality, and on-site electromagnetic interference after actual deployment. The current lack of stress testing methods that reflect actual on-site operating conditions makes it difficult to quantitatively assess the long-term operational reliability of terminals and to support the needs of refined scheduling and differentiated operation and maintenance.

[0006] In summary, how to achieve real-time online monitoring of the communication performance of power energy acquisition terminals and assess their load-bearing capacity under pressure is a technical problem that urgently needs to be solved in the field of power system communication operation and maintenance. Summary of the Invention

[0007] To address the lack of online monitoring methods for the real-time communication performance of power energy acquisition terminals and the lack of a terminal performance evaluation mechanism under stress conditions in existing technologies, this invention provides a real-time monitoring method for the communication performance of power energy terminals based on multi-dimensional stress testing. By establishing a multi-dimensional data acquisition system, constructing a comprehensive performance index model, and combining it with a graded and progressive stress testing mechanism, this method achieves real-time quantitative evaluation and proactive diagnosis of terminal communication performance, thereby providing an effective means of monitoring and early warning of terminal operation status in the context of electricity spot trading.

[0008] To achieve the above objectives, the technical solution adopted by the present invention is as follows:

[0009] A method for real-time monitoring of the communication performance of an electrical energy terminal based on multidimensional stress testing includes the following steps:

[0010] S10. Real-time acquisition of multi-dimensional operating data of the power terminal, the multi-dimensional operating data including downlink communication data, uplink communication data and terminal body data;

[0011] S20. Perform feature extraction and index calculation on the collected multidimensional operation data to obtain a set of communication performance indicators. The set of communication performance indicators includes at least physical layer signal indicators, link layer transmission indicators, network layer connectivity indicators, application layer interaction indicators, and ontology resource indicators.

[0012] S30. Construct a comprehensive performance index model, normalize and weight each index in the communication performance index set, calculate the comprehensive performance index CPI of the terminal, and determine the communication performance status of the terminal based on the comprehensive performance index CPI.

[0013] S40. Execute the stress test steps, trigger the stress test process according to the monitoring requirements, apply simulated load to the terminal according to the preset graded strategy, monitor the changes in the terminal's performance indicators under different stress levels, and generate a performance evaluation report.

[0014] Specifically, in step S10, downlink communication data is acquired by configuring a dedicated sampling chip to capture the differential signal waveform on the RS-485 bus and recording the transmission and reception logs of the link layer protocol frames; uplink communication data is acquired by actively sending ICMP probe packets and listening to the TCP connection status to obtain network layer data and statistically analyzing the application layer data reporting records; terminal body data is acquired by reading CPU utilization, memory usage, core chip temperature, and clock synchronization deviation through the operating system interface.

[0015] Specifically, in step S20, the calculation of physical layer signal indicators includes: calculating the differential voltage amplitude Vpp, signal rise time Tr / signal fall time Tf, and common mode voltage offset Vcm based on sampled waveform data; the calculation of link layer transmission indicators includes: calculating the polling period Tpoll, single read success rate Psucc, response delay distribution P95 / P99 value, and consecutive failure count Nfail based on protocol frame records.

[0016] Specifically, in step S20, the calculation of network layer connectivity indicators includes: calculating round-trip time (RTT), packet loss rate (Ploss), and time jitter based on ICMP probe results; the calculation of application layer interaction indicators includes: calculating statistical packet success rate (Papp), end-to-end latency (Te2e), and concurrent processing capability (Cconc); the calculation of ontology resource indicators includes: calculating the exponentially weighted moving average of CPU utilization, task scheduling delay, temperature rise rate, and clock skew.

[0017] Specifically, the process of constructing the comprehensive performance index model in step S30 is as follows:

[0018] The indicators in the communication performance indicator set are divided into benefit-type indicators and cost-type indicators, and each indicator is mapped to the range of 0 to 100 using the linear normalization method.

[0019] The weight coefficients of each indicator are determined by the analytic hierarchy process (AHP). The weight coefficients include the criterion layer weights and the indicator layer weights. The criterion layer includes downlink communication performance, uplink communication performance and terminal performance.

[0020] The weighted summation method is used to calculate the comprehensive performance index (CPI) based on the normalized values ​​of each indicator and their corresponding weight coefficients.

[0021] Specifically, the method for determining the communication performance status of the terminal based on the Comprehensive Performance Index (CPI) in step S30 is as follows:

[0022] If CPI ≥ 95, the status is considered normal;

[0023] If 90 ≤ CPI < 95, it is considered a general anomaly and the error is logged.

[0024] If 80≤CPI<90, it is considered a serious anomaly, triggering a yellow alert.

[0025] If CPI < 80, it is considered a serious anomaly and a red alert is triggered.

[0026] Specifically, the triggering conditions for triggering the stress test process according to monitoring requirements in step S40 include:

[0027] Performance degradation trigger: Automatically triggered when the comprehensive performance index (CPI) continues to decline for several consecutive monitoring periods and the current value is lower than the preset threshold.

[0028] Manual trigger: Triggered in response to test commands issued by the operations and maintenance terminal;

[0029] Anomaly diagnosis trigger: Triggered when an anomaly is detected and the root cause cannot be determined.

[0030] Specifically, step S40, which involves applying simulated load to the terminal according to a preset hierarchical strategy, includes:

[0031] Normal stress test phase: Simulate the standard communication load of the terminal during daily operation to establish performance baseline data;

[0032] High-stress testing phase: The standard load is increased by a preset multiple to simulate the sudden data reporting peak at critical settlement moments, increase the collection frequency and the number of concurrent requests, and evaluate the stability of the terminal under high load.

[0033] Furthermore, the stress test execution step also includes a recovery monitoring phase:

[0034] After the high-stress test phase is completed, the load will be restored to the standard load of the normal stress test phase.

[0035] Maintain high-frequency data acquisition within a preset time period, monitor changes in performance indicators during the terminal's recovery from high-voltage conditions, and determine whether there is performance degradation or task backlog.

[0036] Specifically, the evaluation of the terminal's stability under high load during the high-stress testing phase includes:

[0037] Set a concurrency gradient, and send a synchronization request at each gradient;

[0038] Record the number of successful responses, average response time, and P95 response time for each gradient.

[0039] When the success rate is lower than the preset success rate threshold or the response time exceeds the preset time threshold, the current concurrency count is recorded as the terminal's critical concurrency capacity.

[0040] Compared with the prior art, the present invention has the following beneficial effects:

[0041] (1) This invention achieves comprehensive real-time monitoring of the communication performance of the power terminal by collecting multi-dimensional operational data of downlink communication, uplink communication, and the terminal itself in real time, and performing feature extraction and index calculation. Compared with the existing technology that focuses on offline detection or single fault diagnosis, this invention can perceive subtle changes such as physical layer signal degradation, link layer transmission delay, and network layer congestion in real time, and promptly detect potential hidden dangers, thereby avoiding the escalation of faults and effectively improving the real-time and comprehensiveness of monitoring.

[0042] (2) This invention constructs a comprehensive performance index model that includes the physical layer, link layer, network layer, application layer, and ontology resources. It uses the analytic hierarchy process (AHP) to determine the weights and performs normalization calculations, quantifying complex communication states into a comprehensive performance index (CPI) ranging from 0 to 100. This quantitative evaluation system enables maintenance personnel to intuitively and accurately grasp the terminal's operating status, solving the problems of vague communication status descriptions and difficulty in horizontal comparisons in existing technologies, and providing scientific data support for refined operation and maintenance.

[0043] (3) By introducing a graded stress test mechanism, this invention can simulate various operating conditions such as normal and high load, especially the sudden high concurrency scenario during the settlement of electricity spot transactions. By actively applying a controllable load, the processing capacity limit and recovery capability of the terminal under stress are evaluated, which solves the problem that the existing technology cannot evaluate the performance of the terminal under extreme conditions. This allows for the early identification of performance bottlenecks and provides preventive measures for communication assurance during peak periods of electricity spot transactions.

[0044] (4) This invention achieves a shift from passively responding to faults to actively preventing faults by establishing a closed-loop process of "real-time monitoring - performance evaluation - stress testing - recovery monitoring". Through the automatic judgment mechanism triggered by stress testing, it can automatically intervene in the test in the early stage of performance degradation, help locate the root cause, greatly shorten the fault response time, reduce the impact of communication interruption on electricity spot trading settlement, and significantly improve the operational reliability of the system. Attached Figure Description

[0045] Figure 1 This is a schematic diagram of the process in an embodiment of the present invention.

[0046] Figure 2 This is a flowchart illustrating the real-time monitoring process in an embodiment of the present invention.

[0047] Figure 3 This is a schematic diagram illustrating the process of constructing the comprehensive performance index in an embodiment of the present invention.

[0048] Figure 4 This is a schematic diagram of the pressure test process in an embodiment of the present invention. Detailed Implementation

[0049] The present invention will be further described below with reference to the accompanying drawings and embodiments. The embodiments of the present invention include, but are not limited to, the following embodiments.

[0050] Example

[0051] like Figures 1 to 4 As shown, this embodiment provides a method for real-time monitoring of the communication performance of power terminals based on multi-dimensional stress testing. This method is applied to the operation and maintenance system of power acquisition terminals in a pilot area for electricity spot trading. The system includes power acquisition terminals (hereinafter referred to as terminals) deployed on-site, a master station system deployed in the operation and maintenance center, and a communication network connecting the two. The method specifically includes the following steps:

[0052] Step S10: Real-time acquisition of multi-dimensional operating data of the power terminal, including downlink communication data, uplink communication data and terminal body data.

[0053] This step aims to establish a comprehensive data perception system, breaking down data silos in traditional monitoring methods. Upon system startup, monitoring configuration parameters are first loaded. These parameters are stored in the terminal's configuration file in JSON format, including indicator threshold parameters, collection cycle parameters, core performance indicators, and alarm rule parameters. The indicator threshold parameters define the normal range, warning threshold, and critical threshold for each monitoring indicator. For example, the normal range for differential voltage amplitude is 2V~6V, the lower warning limit is 1.8V, the upper warning limit is 6.5V, the lower critical limit is 1.5V, and the upper critical limit is 7V. A single read success rate is normally 100%, a warning occurs when one section is not read, and a critical threshold is when no section is read. The collection cycle parameter defines different collection frequencies based on the indicator's rate of change and importance. For example, physical layer / network layer indicators have a 1-minute collection cycle to reflect rapidly changing communication quality, while link layer / application layer indicators have a 5-minute collection cycle; these are statistical indicators and require sample accumulation. Core performance indicators are collected at a 1-minute cycle to monitor resource usage in real time. The alarm rule parameters define the tiered alarm triggering conditions and deduplication strategies. For example, a single violation triggers a level 3 alarm, three consecutive violations trigger a level 2 alarm, and a deteriorating trend (such as a continuous increase exceeding the threshold) triggers a level 1 alarm. The deduplication strategy is that repeated alarms for the same indicator within 30 minutes are only sent once.

[0054] The data acquisition process is executed through the terminal's built-in monitoring module or an external data acquisition unit, and is specifically divided into three parallel acquisition dimensions:

[0055] First, downlink communication data acquisition: This focuses on the RS-485 communication link between the terminal and the energy meter. At the physical layer, a dedicated sampling chip with a sampling rate of at least 20MHz is configured to capture the differential signal waveform on the RS-485 bus. The recording duration must include at least one complete byte frame (approximately 10ms), acquiring the voltage waveforms of lines A and B to ground and the differential voltage. At the link layer, by parsing the terminal communication log, all protocol frame records from the past acquisition cycle (e.g., 5 minutes) are extracted, including the transmit and receive timestamps, address fields, control codes, data lengths, and verification results of DL / T 645 or DL / T 698.45 protocol frames.

[0056] Second, uplink communication data acquisition: This focuses on the Ethernet or wireless communication link between the terminal and the master station. At the network layer, the monitoring module actively sends ICMP Ping packets to the master station's front-end (5 consecutive packets, 100ms interval), recording response times and timeouts; simultaneously, it listens to the terminal operating system's TCP protocol stack statistics to obtain connection status, retransmission counts, etc. At the application layer, it collects statistics on all application layer interactions between the terminal and the master station, including data reporting time, master station confirmation time, and concurrent request queue length.

[0057] Third, terminal data acquisition: CPU usage, memory usage, and task scheduling latency are read through the standard interface of the terminal's embedded operating system (such as the / proc file system or system API); the temperature value of the core communication chip is read through the onboard temperature sensor; and the local clock deviation is calculated by comparing the BeiDou / GPS satellite timing signals.

[0058] All collected data is tagged with a uniform and precise timestamp and temporarily stored in a circular buffer in memory to ensure data time alignment and fast retrieval.

[0059] S20. Perform feature extraction and index calculation on the collected multidimensional operation data to obtain a set of communication performance indicators. The set of communication performance indicators includes at least physical layer signal indicators, link layer transmission indicators, network layer connectivity indicators, application layer interaction indicators, and ontology resource indicators.

[0060] This step utilizes a pre-defined algorithm to perform in-depth processing of the raw data, extracting key quantitative indicators that characterize communication performance. The specific calculation process is as follows:

[0061] In the calculation of physical layer signal indicators, for the differential voltage amplitude Vpp, the algorithm performs peak detection on the captured waveform data, calculates the difference between lines A and B at each sampling point, and takes the average amplitude of 10 consecutive waveform cycles as the current differential voltage amplitude Vpp value. For the signal rise time Tr, the algorithm identifies a complete positive pulse of the differential signal, calculates the voltage thresholds corresponding to 10% and 90% of the pulse amplitude, and records the time difference between crossing these two thresholds as the signal rise time Tr. The signal fall time Tf is calculated similarly. For the common-mode voltage offset Vcm, the average voltage of lines A and B to ground is calculated. For example, in a certain acquisition, the measured Vpp of a terminal's RS-485 interface is 4.5V (normal range 2V-6V), the signal rise time / signal fall time is 0.8μs (normal requirement <1μs), and the common-mode voltage offset is 1V (normal requirement -1V~+3V), indicating that the physical layer signal quality is good.

[0062] In the calculation of link layer transmission indicators, the polling period Tpoll is the time required for the measurement terminal to traverse all connected energy meters once. It is calculated by taking the start and end times of the most recent complete poll and calculating the difference, while excluding interruptions caused by other tasks. The single-read success rate Psucc is calculated by statistically analyzing the number of read attempts and successful responses for a specific energy meter address within 5 minutes, calculating the ratio, and performing independent statistics for each meter. The average success rate for all meters is then calculated. The response latency distribution records the latency of each successful response and calculates statistical characteristics. Specifically, a sliding window of length 100 is maintained, recording the latency values ​​of the most recent 100 successful responses, calculating the average, standard deviation, and P95 and P99 quantiles. P95 is the 95th value obtained by sorting the window latency in ascending order, and P99 is the 99th value obtained by sorting the window latency in ascending order. The consecutive failure count (Nfail) is a record of the number of consecutive reading failures for each meter. It is read from a counter configured on each meter, and the counter is reset to zero when a reading is successful and incremented when a reading fails. For example, a terminal connected to 30 electricity meters has an average reading success rate of 99.8% and a P95 latency of 150ms, indicating high link layer communication efficiency.

[0063] In network layer connectivity metric calculations, average round-trip time (RTT) and packet loss rate (Ploss) are calculated based on ICMP probe results. For example, if five consecutive Ping packets are sent and five responses are received, the average latency is 20ms, jitter is present, and the packet loss rate is 0%. Jitter is calculated as the moving average of the absolute difference between two consecutive successful RTTs. Simultaneously, the increment of TCP retransmission counts is read; a significant increment indicates link congestion.

[0064] In the calculation of application layer interaction metrics, the data packet success rate (Papp) is the proportion of application layer data reports that are confirmed by the master station, determined by the ratio of the number of confirmed reports received to the total number of reported data packets. End-to-end latency (Te2e) is the total time elapsed from data generation to master station confirmation, calculated based on the difference between the generation time and the confirmation time for each reported record. The key evaluation focuses on concurrent processing capability (Cconc). The system performs concurrent tests during idle periods (e.g., 2 AM), with the concurrency level set to increase from 1 to 50. When the concurrency level reaches 30, the test shows that the P95 response time exceeds the 1-second threshold, with a success rate of 97%. Therefore, the critical concurrency capability of this terminal is determined to be around 25.

[0065] The entity resource indicators are directly calculated based on the terminal entity data collection data.

[0066] S30. Construct a comprehensive performance index model, normalize and weight each index in the communication performance index set, calculate the comprehensive performance index CPI of the terminal, and determine the communication performance status of the terminal based on the comprehensive performance index CPI.

[0067] To intuitively evaluate the overall performance of the terminal, this embodiment constructs a comprehensive performance index model, which mainly consists of three sub-steps: index normalization, weight allocation, and index calculation. The specific process is as follows:

[0068] First, the metrics are normalized. The metrics in the communication performance metric set are divided into benefit-based metrics and cost-based metrics, and each metric is mapped to a range of 0 to 100 using a linear normalization method. Benefit-based metrics include command success rate, data packet success rate, concurrent processing capability, throughput, and time synchronization success rate, etc. Higher values ​​indicate better performance; for example, a higher success rate is better. The following formula is used for normalization:

[0069]

[0070] In the above formula, X is the measured value of the corresponding indicator, and Xmin and Xmax are the historical minimum and historical maximum values ​​(or theoretical thresholds) of the corresponding indicator, respectively.

[0071] Cost-related metrics include response latency, round-trip latency, packet loss rate, TCP retransmission count, CPU utilization, memory utilization, chip temperature, temperature rise rate, clock skew, consecutive failure count, signal rise / fall time, and common-mode voltage offset. Lower values ​​generally indicate better performance; for example, lower latency is better. These metrics are normalized using the following formula:

[0072]

[0073] In the above formula, X is the measured value of the corresponding indicator, and Xmin and Xmax are the historical minimum and historical maximum values ​​(or theoretical thresholds) of the corresponding indicator, respectively. For example, if the current packet loss rate of a terminal is 1% and the historical maximum packet loss rate threshold is 10%, then its normalized indicator value Xnorm = (10-1) / (10-0) × 100 = 90 points.

[0074] The closer the normalized index value Xnorm is to 100, the better the performance of that index dimension; the closer it is to 0, the more severe the performance degradation.

[0075] Secondly, the Analytic Hierarchy Process (AHP) is used to determine the weight coefficients of each indicator. These weight coefficients include criterion-level weights and indicator-level weights. The criterion-level weights include Downlink Communication Performance (DCP), Uplink Communication Performance (UPC), and Terminal Performance (TP). Downlink Communication Performance reflects the RS-485 bus communication quality between the terminal and the electricity meter; Uplink Communication Performance reflects the Ethernet / wireless communication quality between the terminal and the master station; and Terminal Performance reflects the terminal's own resource consumption, thermal characteristics, and clock synchronization status. Based on the actual communication performance requirements of electricity spot trading, this embodiment allocates the weights of these three criterion-level weights as follows: Downlink Communication Performance accounts for 40% of the weight, Uplink Communication Performance accounts for 40%, and Terminal Performance accounts for 20%.

[0076] The indicator layer weights are the weights of multiple sub-dimensional indicators further subdivided within each criterion layer. The rationality of the weight allocation is ensured through expert scoring and historical data fitting verification. This embodiment uses a system weight coefficient as an example: Within the downlink communication performance criterion layer, physical layer signal quality accounts for 15% of the weight, including differential voltage amplitude Vpp (5%), signal rise time Tr / signal fall time Tf (4%), common-mode voltage offset Vcm (3%), and noise amplitude Vn (3%). Link layer communication efficiency accounts for 25% of the weight, including single-pass read success rate Psucc (10%), polling period Tpoll (5%), response delay distribution P95 (7%), and consecutive failure count Nfail (3%). Within the uplink communication performance criterion layer, network layer connectivity accounts for 15% of the weight, including round-trip time RTT (5%), packet loss rate Ploss (5%), and so on. The weighting is as follows: 5% for latency jitter (3%), 2% for TCP retransmission count (Nretrans), 25% for application layer transmission efficiency (including packet success rate Papp, 10%), 7% for end-to-end latency Te2e, 5% for concurrent processing capability Cconc, and 3% for throughput. Within the terminal performance criteria layer, computing resources account for 10% of the weighting, including CPU utilization (5%), task scheduling latency (5%), thermal performance (5%), including core chip temperature (3%), temperature rise rate (2%), and clock synchronization (5%), including local clock deviation (3%) and time synchronization success rate (2%).

[0077] Finally, the weighted summation method is used to calculate the Comprehensive Performance Index (CPI) based on the normalized values ​​of each indicator and their corresponding weight coefficients. The formula is as follows:

[0078]

[0079] In the above formula, W i Let X be the global weight of the i-th indicator. i norm Here, is the normalized value of the i-th indicator, and n is the total number of indicators, for example, n=19 in this embodiment. The CPI ranges from 0 to 100, with higher values ​​indicating better overall terminal communication performance. For example, a terminal's calculated scores are as follows: the weighted score for each indicator within the downlink communication performance criterion layer is 38 points, the weighted score for each indicator within the uplink communication performance criterion layer is 36 points, the weighted score for each indicator within the terminal body performance criterion layer is 18 points, and the final CPI is 92 points.

[0080] The Comprehensive Performance Index (CPI) is divided into the following levels: 95-100 is excellent (green), indicating good communication performance; 90-95 is good (blue), indicating normal communication performance; 80-90 is poor (orange), indicating a decline in communication performance; and 0-80 is dangerous (red), indicating a serious communication anomaly.

[0081] The specific method for determining the communication performance status of a terminal based on the Comprehensive Performance Index (CPI) is as follows:

[0082] If CPI ≥ 95, the condition is considered normal, and routine monitoring should continue.

[0083] If 90 ≤ CPI < 95, it is judged as a general anomaly, logged, and marked as a concern.

[0084] If 80≤CPI<90, it is considered a serious anomaly, triggering a yellow alarm, and on-site staff must be notified.

[0085] If CPI < 80, it is considered a serious anomaly, triggering a red alert, and requires immediate action.

[0086] For example, if the terminal's CPI score is 92, it is judged as a general anomaly, and the system logs and marks it as a concern.

[0087] S40. Execute the stress test steps, trigger the stress test process according to the monitoring requirements, apply simulated load to the terminal according to the preset graded strategy, monitor the changes in the terminal's performance indicators under different stress levels, and generate a performance evaluation report.

[0088] The stress test in this step aims to simulate various communication stress scenarios that the terminal may encounter in actual operation by actively applying a controllable load, and to evaluate the terminal's performance, stability limits, and fault recovery capabilities under different load levels. The stress test adopts a tiered and progressive mode, dynamically triggered according to the terminal's current health status and operational needs, and forms a closed-loop linkage with regular real-time monitoring.

[0089] The stress test is initiated by a system-defined trigger mechanism based on preset conditions. The system uses "OR" logic for comprehensive decision-making. The trigger conditions for the stress test process include:

[0090] Performance degradation trigger: Automatically triggered when the comprehensive performance index (CPI) continues to decline for several consecutive monitoring periods and the current value is below the preset threshold; for example, automatic trigger when the CPI continues to decline for 3 consecutive monitoring periods and the current value is below 90 points. This condition aims to proactively access the performance degradation process and locate potential bottlenecks.

[0091] Manual trigger: Triggered in response to test commands issued by the operation and maintenance terminal; for example, operation and maintenance personnel can actively issue stress test commands to a specified terminal or group of terminals through the main station system or mobile operation and maintenance terminal. This condition supports planned inspections, fault reproduction verification, and capacity assessment before the launch of new services.

[0092] Anomaly diagnosis trigger: This is triggered when an anomaly is detected and the root cause cannot be determined. Specifically, when the anomaly detection module identifies an anomaly but cannot determine the root cause through knowledge base matching, a stress test is triggered to reproduce the anomaly scenario and assist in root cause localization.

[0093] In stress testing, this invention employs a pre-defined tiered stress testing strategy, dividing the stress test into three stages based on the testing objectives and risk control requirements. The load parameters, duration, and data acquisition frequency for each stage are dynamically adjusted to ensure that performance is fully evaluated while avoiding unacceptable impacts on on-site operations.

[0094] The first phase, normal stress testing, aims to establish the performance baseline of the terminal and obtain its reference standard under standard operating conditions, providing a benchmark for subsequent stress testing.

[0095] Load setting: Simulates the standard communication load of a terminal during daily operation.

[0096] Settings: Reading cycle 15 minutes, data reporting cycle 15 minutes, concurrent reading task not exceeding 10, meter scale simulating 30 meters, duration is continuous operation synchronized with the regular monitoring cycle, and baseline data is formed daily. Collection frequency: Physical layer / network layer indicators are collected at a 1-minute cycle (same as regular monitoring), and link layer / application layer indicators are collected at a 5-minute cycle (same as regular monitoring).

[0097] Test objective: To obtain reference values ​​for terminal performance indicators under normal load, including single-pass read success rate, polling cycle, response latency distribution, end-to-end latency, etc., and to establish dynamic baselines for each indicator.

[0098] The second phase, high-stress testing, simulates the surge in data reporting during critical settlement times in electricity spot trading, such as the hour and half-hour, to evaluate the performance stability of the terminal under high load.

[0099] Load setting: Increase the standard load by 2 to 4 times to simulate daily peak pressure.

[0100] Settings: The reading cycle is compressed to 5 minutes, the concurrency is increased to 20-50, and temporary data collection tasks are added, such as high-frequency data collection from simulated meters every 3 minutes. The duration is set to 20-30 minutes, covering 1-2 complete settlement cycles to ensure sufficient sample collection. To capture transient response characteristics, the collection frequency is increased to a 10-second cycle, including instruction-level response time recording, high-frequency sampling of CPU / memory usage, and real-time monitoring of channel load.

[0101] Test Objective: To evaluate the success rate, response time, and resource usage changes of the terminal under high pressure; to verify the mutual impact of concurrent multi-task execution (e.g., whether high-frequency data collection affects daily frozen tasks); and to detect software defects such as memory leaks and task backlogs. For example, during the high-pressure test, it was found that the terminal's CPU utilization surged to 95% in the 5th minute after the high pressure began, the data reading success rate dropped to 92%, and the P95 response time increased to 2 seconds. The system recorded this phenomenon and determined that the terminal had a resource bottleneck under high-concurrency scenarios.

[0102] The third stage, recovery monitoring, is executed automatically after the stress test ends. It is used to observe the performance of the terminal after it recovers from a high-pressure state to a normal load state, and to evaluate its recovery capability and whether there is any performance degradation.

[0103] Load setting: Restore to the standard load of normal stress testing.

[0104] Settings: Duration 20-30 minutes, with the first 10 minutes maintaining a high-frequency acquisition frequency of 10-second intervals to observe the transient recovery process, and the next 10-20 minutes resuming the normal acquisition frequency (1 minute / 5 minutes) to confirm long-term stability.

[0105] Test objective: To observe the terminal's ability to recover from high-pressure conditions, detect any performance degradation, and verify the fault self-recovery mechanism. For example, during the test, it was found that after the load decreased, the CPU utilization rate quickly dropped to 20%, but the copying success rate remained at around 98% for the next 2 minutes, suggesting that there might be a backlog of tasks being processed. After 10 minutes, all indicators returned to normal.

[0106] After the test, the system generated a detailed performance evaluation report, indicating that the terminal's critical concurrency capacity was approximately 25. It recommended controlling the number of concurrent tasks during peak electricity spot trading periods and, alternatively, upgrading the hardware and software. Maintenance personnel can then adjust the terminal's data reporting strategy accordingly to mitigate the potential risk of transaction data upload failures.

[0107] Therefore, the method of the present invention can achieve in-depth perception and quantitative evaluation of the communication performance of power terminals. In particular, through the stress test, it can proactively expose the performance shortcomings of terminals under extreme conditions, providing strong technical support for the reliable operation of electricity spot trading.

[0108] The above embodiments are merely preferred embodiments of the present invention and are not intended to limit the scope of protection of the present invention. Any changes made based on the design principles of the present invention, or any non-creative modifications made thereon, shall fall within the scope of protection of the present invention.

Claims

1. A method for real-time monitoring of the communication performance of an electrical energy terminal based on multi-dimensional stress testing, characterized in that, Includes the following steps: S10. Real-time acquisition of multi-dimensional operating data of the power terminal, the multi-dimensional operating data including downlink communication data, uplink communication data and terminal body data; S20. Perform feature extraction and index calculation on the collected multidimensional operation data to obtain a set of communication performance indicators. The set of communication performance indicators includes at least physical layer signal indicators, link layer transmission indicators, network layer connectivity indicators, application layer interaction indicators, and ontology resource indicators. S30. Construct a comprehensive performance index model, normalize and weight each index in the communication performance index set, calculate the comprehensive performance index CPI of the terminal, and determine the communication performance status of the terminal based on the comprehensive performance index CPI. S40. Execute the stress test steps, trigger the stress test process according to the monitoring requirements, apply simulated load to the terminal according to the preset graded strategy, monitor the changes in the terminal's performance indicators under different stress levels, and generate a performance evaluation report.

2. The method for real-time monitoring of the communication performance of an electrical energy terminal based on multi-dimensional stress testing according to claim 1, characterized in that, In step S10, downlink communication data is acquired by configuring a dedicated sampling chip to capture the differential signal waveform on the RS-485 bus and recording the transmission and reception logs of the link layer protocol frames; uplink communication data is acquired by actively sending ICMP probe packets and listening to the TCP connection status to obtain network layer data and statistically analyzing the application layer data reporting records; terminal body data is acquired by reading CPU utilization, memory usage, core chip temperature, and clock synchronization deviation through the operating system interface.

3. The method for real-time monitoring of the communication performance of an electrical energy terminal based on multi-dimensional stress testing according to claim 1, characterized in that, In step S20, the calculation of physical layer signal indicators includes: calculating differential voltage amplitude Vpp, signal rise time Tr / signal fall time Tf and common mode voltage offset Vcm based on sampled waveform data; the calculation of link layer transmission indicators includes: calculating polling period Tpoll, single copy success rate Psucc, response delay distribution P95 / P99 value and consecutive failure count Nfail based on protocol frame records.

4. The method for real-time monitoring of the communication performance of an electrical energy terminal based on multi-dimensional stress testing according to claim 1, characterized in that, In step S20, the calculation of network layer connectivity indicators includes: calculating round-trip time (RTT), packet loss rate (Ploss), and time jitter based on ICMP probe results; the calculation of application layer interaction indicators includes: calculating statistical packet success rate (Papp), end-to-end latency (Te2e), and concurrent processing capability (Cconc); the calculation of ontology resource indicators includes: calculating the exponentially weighted moving average of CPU utilization, task scheduling delay, temperature rise rate, and clock skew.

5. The method for real-time monitoring of the communication performance of an electrical energy terminal based on multi-dimensional stress testing according to claim 1, characterized in that, The specific process of constructing the comprehensive performance index model in step S30 is as follows: The indicators in the communication performance indicator set are divided into benefit-type indicators and cost-type indicators, and each indicator is mapped to the range of 0 to 100 using the linear normalization method. The weight coefficients of each indicator are determined by the analytic hierarchy process (AHP). The weight coefficients include the criterion layer weights and the indicator layer weights. The criterion layer includes downlink communication performance, uplink communication performance and terminal performance. The weighted summation method is used to calculate the comprehensive performance index (CPI) based on the normalized values ​​of each indicator and their corresponding weight coefficients.

6. The method for real-time monitoring of the communication performance of an electrical energy terminal based on multi-dimensional stress testing according to claim 5, characterized in that, The specific method for determining the communication performance status of the terminal based on the Comprehensive Performance Index (CPI) in step S30 is as follows: If CPI ≥ 95, the status is considered normal; If 90 ≤ CPI < 95, it is considered a general anomaly and the error is logged. If 80≤CPI<90, it is considered a serious anomaly, triggering a yellow alert. If CPI < 80, it is considered a serious anomaly and a red alert is triggered.

7. The method for real-time monitoring of the communication performance of an electric energy terminal based on multi-dimensional stress testing according to claim 1, characterized in that, The triggering conditions for initiating the stress test process according to monitoring requirements in step S40 include: Performance degradation trigger: Automatically triggered when the comprehensive performance index (CPI) continues to decline for several consecutive monitoring periods and the current value is lower than the preset threshold. Manual trigger: Triggered in response to test commands issued by the operations and maintenance terminal; Anomaly diagnosis trigger: Triggered when an anomaly is detected and the root cause cannot be determined.

8. The method for real-time monitoring of the communication performance of an electrical energy terminal based on multi-dimensional stress testing according to claim 1, characterized in that, The step S40, which involves applying simulated load to the terminal according to a preset hierarchical strategy, specifically includes: Normal stress test phase: Simulate the standard communication load of the terminal during daily operation to establish performance baseline data; High-stress testing phase: The standard load is increased by a preset multiple to simulate the sudden data reporting peak at critical settlement moments, increase the collection frequency and the number of concurrent requests, and evaluate the stability of the terminal under high load.

9. The method for real-time monitoring of the communication performance of an electrical energy terminal based on multi-dimensional stress testing according to claim 8, characterized in that, The stress test execution step also includes a recovery monitoring phase: After the high-stress test phase is completed, the load will be restored to the standard load of the normal stress test phase. Maintain high-frequency data acquisition within a preset time period, monitor changes in performance indicators during the terminal's recovery from high-voltage conditions, and determine whether there is performance degradation or task backlog.

10. The method for real-time monitoring of the communication performance of an electrical energy terminal based on multi-dimensional stress testing according to claim 8, characterized in that, The evaluation of the terminal's stability under high load during the high-stress testing phase specifically includes: Set a concurrency gradient, and send a synchronization request at each gradient; Record the number of successful responses, average response time, and P95 response time for each gradient. When the success rate is lower than the preset success rate threshold or the response time exceeds the preset time threshold, the current concurrency count is recorded as the terminal's critical concurrency capacity.

Images