A two-wire power supply control system fault diagnosis method and system

Abstract

The application discloses a kind of two-wire system power supply control system fault diagnosis method and system, belong to power supply control technical field, including physical layer, perception layer, node data acquisition layer and decision-making execution layer.Physical layer uses prefabricated quick connection cable and T type intelligent joint, each cable solidifies standard resistance parameter;Perception layer is only set current sensor at power supply end, and each node is collected voltage by time synchronization;Node data acquisition layer obtains theoretical electrical quantity;Decision-making execution layer uses the step-by-step investigation strategy from root to leaf, calculates the current residual of each cable by voltage difference reconstruction algorithm, locates the electric leakage fault by node current residual analysis, and realizes full working condition coverage by using adaptive monitoring strategy.The system also integrates the anti-theft electricity and illegal access monitoring module to realize the accurate identification and positioning of illegal access.The application realizes high resistance fault active early warning, meter-level accurate positioning and anti-theft electricity function, and is suitable for garden landscape, agricultural greenhouse and other scenes.

Description

Technical Field

[0001] This invention belongs to the field of power supply control technology, and in particular relates to a fault diagnosis method and system for a two-wire power supply control system. Background Technology

[0002] In two-wire low-voltage power supply control systems for landscape lighting, agricultural greenhouses, and municipal lighting, line faults (such as leakage, short circuits, poor contact, and cable aging) occur frequently due to the complex environment in which the lines are laid (humidity, soil corrosion, rodent damage, and human-caused construction damage). Furthermore, the fault points are often concealed, making troubleshooting difficult. Existing technologies have the following main drawbacks:

[0003] Traditional methods rely on multimeters to measure segment by segment and manual troubleshooting, which is inefficient, cannot achieve real-time monitoring, and has a lag in fault detection, often leading to the expansion of faults or even causing safety accidents.

[0004] Existing intelligent lighting systems often use current sensors installed at the end of each circuit or at each branch node, which significantly increases hardware costs. Furthermore, in decentralized scenarios such as gardens, wiring is complex and installation and maintenance are difficult, which is not conducive to large-scale promotion.

[0005] Traditional residual current devices (RCDs) are mainly designed for tripping due to high current leakage. They often fail to provide timely warnings for "high impedance faults" (which generate a lot of heat but do not show significant current changes) caused by factors such as joint oxidation, loose connections, and cable corrosion, posing a fire hazard.

[0006] The existing system does not make full use of the characteristic of "prefabricated cables" in garden line construction (that is, the cable length, resistance and other parameters are determined at the factory), which leads to difficulties in topology identification, cumbersome parameter matching and poor project consistency after system installation.

[0007] For measuring small resistance in short-distance lines, the direct measurement method has poor linearity and low accuracy when the current change is small in the early stage of leakage. At the same time, when the line is unloaded or the current is too small, the traditional voltage-current detection method is completely ineffective.

[0008] In scenarios such as garden landscape lighting, agricultural greenhouses, and municipal lighting, the lines are scattered and unattended, making it difficult to detect illegal access (electricity theft) in a timely manner. Traditional methods cannot distinguish between normal load changes and illegal access, and there is a lack of effective technical means.

[0009] Therefore, there is an urgent need for an intelligent fault diagnosis system that is low-cost, high-precision, covers all working conditions, and has the ability to accurately locate faults and prevent electricity theft. Summary of the Invention

[0010] To address the shortcomings of existing technologies, this invention provides a fault diagnosis method and system for a two-wire power supply control system, thus solving the aforementioned problems.

[0011] To achieve the above objectives, the present invention provides the following technical solution: a fault diagnosis system for a two-wire power supply control system, comprising:

[0012] Physical layer: Standardized prefabricated quick-connect cables and T-type smart connectors are used to construct a tree-like or daisy-chain topology; each prefabricated cable segment has its length, standard resistance value and unique electronic tag fixed at the factory;

[0013] The sensing layer includes a main control unit installed at the power supply end and voltage monitoring terminals distributed at each T-type smart connector. The main control unit has a total current sampling function, and each voltage monitoring terminal collects the voltage value of the corresponding node. All monitoring terminals have a time synchronization mechanism. In the sensing layer, only the power supply end is equipped with a current sensor. The load current is measured by the current sampling circuit built into the load device and reported to the main control unit through the communication network. No current sensor is installed at the T-type node.

[0014] Node data acquisition layer: Real-time acquisition of theoretical voltage and theoretical current values ​​for each node;

[0015] Decision execution layer: Fault diagnosis and location are performed using voltage difference reconstruction algorithm and node current residual analysis. The decision execution layer adopts a step-by-step investigation strategy from root to leaf, taking the estimated current obtained by voltage difference reconstruction of upstream nodes as the known input current of downstream nodes, and calculating the residual between the theoretical current and the measured estimated current of each branch step by step to achieve accurate fault location.

[0016] Based on the above technical solutions, the present invention also provides the following optional technical solutions:

[0017] A further technical solution: The voltage difference reconstruction algorithm includes: obtaining the voltage of two adjacent nodes. , ; Calculate the voltage difference ;

[0018] According to the standard resistance of this prefabricated cable section Calculate the measured and estimated current Obtain the theoretical current of this cable segment. ; Calculate the current residual ;like Determine the health of the line; if This indicates an abnormally low current; if The system determined that there was an abnormally high current; among them, The threshold for judging current residual is set.

[0019] Further technical solution: The node current residual analysis includes: defining the first... The inflow current at each T-node is The outflow current is The load current is ; Calculate the node current residual ;like If the leakage current exceeds the preset leakage current threshold, it is determined that a leakage current or grounding fault has occurred at the node or its downstream.

[0020] Further technical solutions include adaptive monitoring strategies.

[0021] When the total current at the power supply terminal exceeds the preset threshold, the voltage drop residual analysis mode is entered.

[0022] When the total current at the power supply end is less than the preset threshold, switch to the level consistency monitoring mode;

[0023] In no-load mode, the system control power supply terminal actively injects constant current test pulses, uses the voltage difference during the pulse to calculate the actual line resistance, and compares it with the standard resistance to determine the line health.

[0024] Further technical solutions: The step-by-step investigation strategy includes:

[0025] Main line diagnosis: Calculating the estimated current of the main line based on voltage difference reconfiguration Calculate the residual of node A. If the threshold is exceeded, the main line is considered to be faulty;

[0026] Branch diagnosis: As the input current of node A, calculate the residual between the theoretical current and the measured estimated current of the downstream branch, and determine the health status of the branch.

[0027] Multi-level expansion: Repeated branch diagnostic logic, proceeding downwards level by level until all end nodes are reached;

[0028] in, This is the power supply voltage. For the voltage at node A, The standard resistor for prefabricated cables in the main trunk line. To calculate the current of the main line based on voltage difference reconstruction, This represents the total current measured at the power supply terminal. Let the current residual at node A be... Let be the current flowing into node A.

[0029] Further technical solutions include an anti-theft and unauthorized access monitoring module; this module, based on voltage difference reconstruction and node current residual analysis, determines whether unauthorized load access exists according to the following steps:

[0030] Step 1: Calculate the outlier confidence levels for the following four factors: Factor A: Node power imbalance confidence level Factor B: Impedance characteristic matching confidence level Factor C: Confidence level of historical current baseline deviation Factor E: Confidence level of timing correlation of load switch events ;

[0031] Step 2: Calculate the overall confidence level using a multi-factor fusion formula: ,in, The overall confidence level is set and the value range is [0,1]. As a factor index, and taking A, B, C, E, For the first The preset weights of each factor and satisfy the following conditions: ; For the first The outlier confidence level of each factor, with a value range of [0,1];

[0032] Step 3: Assemble the overall confidence level With the first preset threshold and the second preset threshold Comparison: If If it determines that there is electricity theft or illegal access, an alarm will be triggered; if If marked as suspected, an active pulse detection will be automatically triggered for further confirmation; if The result is normal.

[0033] Further technical solution: The anti-electricity theft and unauthorized access monitoring module also includes a confidence correction layer. This layer dynamically corrects the overall confidence level based on one or more environmental and historical behavioral factors to obtain a final confidence level, which is then used to replace the overall confidence level for threshold comparison.

[0034] Factor F1: Factors affecting soil moisture: ,in, Soil moisture influencing factors To preset the sensitivity coefficient, This refers to the measured or estimated volumetric water content of the soil. For reference humidity, To take the positive part of the function, that is, when the part inside the parentheses is negative, it takes the value 0;

[0035] Factor F2: Circuit thermal aging accumulation factor: , ,in, This is the cumulative factor for thermal aging of the circuit. The aging index, To calculate the total power-on time, For cable design life, The temperature acceleration coefficient, The average operating temperature, For reference temperature, This is the aging inhibition coefficient;

[0036] Factor F3: Day / Night / Seasonal Pattern Factor: ,in, For day / night / seasonal pattern factors, The preset suppression coefficient, This is an indicator function that takes the value 1 if any of the following conditions are met, and 0 otherwise: the current time is during the late night period;

[0037] The current time is within a historically low load period for this time of year;

[0038] Factor F4: Consistency factor in the behavior of neighboring nodes. ,in, This serves as a consistency factor for the behavior of adjacent nodes. For the preset enhancement coefficient, This represents the standard deviation of the anomaly confidence scores of each child node under the same parent node. The standard deviation of the confidence level of historical anomalies across the entire network;

[0039] Factor F5: Load switch fatigue index: , ,in, For load switch fatigue factor, The fatigue index. This represents the average number of times the device is switched on and off per day. To design the rated number of switching cycles, This is the fatigue inhibition coefficient;

[0040] Final confidence level correction formula: ,in, The final confidence level is defined with a value range of [0,1]. To assess the overall confidence level, For factor index, No. One correction factor, For the first The enable flag for each factor is either 0 or 1, where 0 indicates that the factor does not participate in the correction and 1 indicates that it does.

[0041] by With the first preset threshold and the second preset threshold Make a comparative decision.

[0042] Further technical solutions: For each factor , =1, 2, 3, 4, 5 correspond to F1~F5 respectively, and their activation flags are... Determined according to the following rules:

[0043] Rule 1: If the system has the data source required for the factor, then The initial value can be set to 1, otherwise it should be set to 0.

[0044] Rule 2: System administrators can manually configure the activation status of each factor through the host computer software, overriding the default value;

[0045] Rule 3: The system can automatically adjust based on historical false alarm rates: If enabling a certain factor leads to a decrease in the overall confidence level... If the consistency with the manual review result is below the threshold, the factor will be automatically disabled and logged.

[0046] Rule 4: For factors lacking real-time data, temporarily set... =0, will be restored after the condition is met.

[0047] A fault diagnosis method for a two-wire power supply control system is implemented using the system described in any one of claims 1-8.

[0048] This invention provides a fault diagnosis method and system for a two-wire power supply control system, which has the following advantages compared with the prior art:

[0049] 1. This invention utilizes the principles of "parameter prior" and "voltage difference reconstruction". Only one current sensor needs to be set at the power supply end. Each T-node only needs voltage sampling. There is no need to install an independent current sensor at each branch node, which greatly reduces hardware costs and construction complexity, and replaces hardware stacking with software algorithms.

[0050] 2. This invention, through current residual analysis and node current residual analysis, can detect early hidden dangers of high impedance such as "loose connection", "oxidation" and "insulation aging" in advance, realizing the transformation from "passive protection" to "active early warning" and effectively preventing electrical fires.

[0051] 3. This invention adopts a step-by-step troubleshooting strategy from root to leaf, combined with voltage difference reconstruction and node current residual analysis, which can improve the fault location accuracy to a specific section of prefabricated cable or a specific T-connector. Maintenance personnel can achieve modular and rapid replacement, which greatly reduces operation and maintenance costs.

[0052] 4. This invention solves the problem of complete failure of traditional methods under no-load and light-load conditions by adopting adaptive monitoring strategies (high current mode, low current / no-load mode, active pulse injection mode), and realizes reliable diagnosis and preventive detection of the system under all operating conditions.

[0053] 5. This invention integrates five factors, including power imbalance, impedance characteristic matching, historical current baseline deviation, active pulse reflection characteristics, and load switch event timing correlation, combined with a confidence correction layer (soil moisture, line thermal aging, day and night seasonal patterns, adjacent node consistency, and load switch fatigue index), to identify illegal access behavior with high accuracy and locate the electricity theft point. It is suitable for unattended garden scenarios. Attached Figure Description

[0054] Figure 1 This is a schematic diagram of the process of the present invention. Detailed Implementation

[0055] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the invention.

[0056] The specific implementation of the present invention will be described in detail below with reference to specific embodiments.

[0057] Please see Figure 1 A fault diagnosis system for a two-wire power supply control system, provided in one embodiment of the present invention, includes:

[0058] Physical layer: Standardized prefabricated quick-connect cables and T-type smart connectors are used to construct a tree-like or daisy-chain topology; each prefabricated cable segment has its length, standard resistance value and unique electronic tag fixed at the factory;

[0059] The sensing layer includes a main control unit installed at the power supply end and voltage monitoring terminals distributed at each T-type smart connector. The main control unit has a total current sampling function, and each voltage monitoring terminal collects the voltage value of the corresponding node. All monitoring terminals have a time synchronization mechanism. In the sensing layer, only the power supply end is equipped with a current sensor. The load current is measured by the current sampling circuit built into the load device and reported to the main control unit through the communication network. No current sensor is installed at the T-type node.

[0060] Node data acquisition layer: Real-time acquisition of theoretical voltage and theoretical current values ​​for each node;

[0061] Decision execution layer: Fault diagnosis and location are performed using voltage difference reconstruction algorithm and node current residual analysis. The decision execution layer adopts a step-by-step investigation strategy from root to leaf, taking the estimated current obtained by voltage difference reconstruction of upstream nodes as the known input current of downstream nodes, and calculating the residual between the theoretical current and the measured estimated current of each branch step by step to achieve accurate fault location.

[0062] The following example will provide a more detailed explanation of the above technical solution:

[0063] Imagine a two-wire power supply control system for a large agricultural greenhouse. This system is deployed to monitor and diagnose line faults. The greenhouse's power supply network uses a tree topology, consisting of multiple standardized prefabricated quick-connect cables and T-type smart connectors. Each cable is pre-programmed with its precise length, standard resistance value, and a unique electronic tag at the factory. This information is automatically identified and entered into the main control unit's database during system installation.

[0064] During system operation, the main control unit continuously collects the total current at the power supply end. Simultaneously, voltage monitoring terminals distributed at each T-type smart connector synchronously collect the voltage values ​​of their respective nodes under a precise time synchronization mechanism. For example, when the main control unit issues a synchronization command, all voltage monitoring terminals simultaneously sample voltage after a preset microsecond delay, ensuring data timing consistency. In this system, no current sensors are installed at the T-type nodes; instead, each load device (e.g., irrigation pump, fan, supplemental lighting, etc.) has its own built-in current sampling circuit to measure its own load current and reports this data to the main control unit in real time via the communication network.

[0065] The node data acquisition layer calculates and obtains the theoretical voltage and current values ​​of each node in a healthy state in real time based on the preset topology, cable parameters, and known load distribution. These theoretical values ​​constitute the benchmark for the system to perform fault diagnosis.

[0066] When the system detects a potential anomaly, the decision execution layer initiates a fault diagnosis process. This layer first utilizes a voltage difference reconstruction algorithm. For example, for any two adjacent nodes A and B on the main line, the main control unit obtains the voltage values ​​collected by their voltage monitoring terminals. and According to the standard resistance value fixed at the factory for this section of prefabricated cable. The system calculates the measured and estimated current of this cable segment. Subsequently, the estimated current... Theoretical current provided by the node data acquisition layer The current residual is calculated by comparing the results.

[0067] Furthermore, the decision-making execution layer employs a root-to-leaf hierarchical troubleshooting strategy. This strategy begins at the power supply end, first diagnosing the main line. For example, the main control unit uses the total current at the power supply end as the input current of the main line, and combines it with the estimated current of the main line calculated by the voltage difference reconstruction algorithm to perform residual analysis. If an anomaly is found in the main line, the system will immediately issue an early warning. If the main line is healthy, its estimated current value is used as the known input current of the first downstream T-node.

[0068] Next, the system performs node current residual analysis on the T-type node. For example, for node C, the system obtains the estimated current flowing into that node. And the estimated current flowing from that node to each branch. , Meanwhile, it acquires the load current reported by the load devices connected to this node. The system calculates the node current residuals. If the residual exceeds a preset threshold, it is determined that there is an anomaly such as leakage or illegal access in the node or its downstream.

[0069] This system achieves precise fault location through voltage difference reconstruction algorithms and node current residual analysis, combined with a root-to-leaf hierarchical troubleshooting strategy. For example, in the aforementioned agricultural greenhouse example, the system can pinpoint the fault to a specific line segment or T-node, rather than simply indicating a large area. This precise location capability significantly improves the efficiency of fault diagnosis and reduces the workload of maintenance personnel. Simultaneously, node current residual analysis also provides a basis for monitoring unauthorized access. By identifying abnormal current imbalances at nodes, the system can distinguish between normal load changes and unauthorized load access, compensating for the shortcomings of existing technologies in preventing electricity theft.

[0070] Preferably, the voltage difference reconstruction algorithm includes: obtaining the voltage of two adjacent nodes. , ; Calculate the voltage difference According to the standard resistance of this prefabricated cable section Calculate the measured and estimated current Obtain the theoretical current of this cable segment. ;

[0071] Calculate current residual ;like Determine the health of the line; if This indicates an abnormally low current; if The system detected an abnormally high current (possibly due to insulation damage, partial short circuit, or leakage); among which, The threshold for judging current residual is set.

[0072] Among them, the voltage of two adjacent nodes is obtained. , This is the basic data for calculating the voltage difference. This voltage value can be collected in real time by a voltage monitoring terminal installed at the T-type smart connector and then acquired by the main control unit via a communication bus; alternatively, each voltage monitoring terminal can collect the data locally and then report it to the main control unit through a time synchronization mechanism. (Calculate the voltage difference.) The aim is to quantify the potential drop between adjacent nodes to reflect energy loss along the cable segment. This calculation can be performed by the main control unit after receiving the voltage values ​​of the two nodes, or it can be performed locally by the microcontroller inside the T-connector and the result reported. This is based on the standard resistance of the prefabricated cable segment. Calculate the measured and estimated current The purpose is to deduce the actual current flowing through a section of cable based on Ohm's law, using the measured voltage difference and the known standard resistance of the cable. The main control unit can query the corresponding cable from the system database. The calculation is then performed, or stored in the electronic tag of the cable. The voltage is read and calculated by the voltage monitoring terminal to obtain the theoretical current of this cable segment. The aim is to provide a benchmark value for comparison with the measured and estimated current to determine if any anomalies exist. This theoretical current can be estimated based on the rated current, operating status, or historical data of downstream load equipment, or it can be obtained in real time by comprehensively considering the communication-reported current of the load equipment and the system's preset load model through the node data acquisition layer. Calculate the current residual. The aim is to quantify the deviation between theoretical current and measured estimated current, which is a key indicator for judging the health of the line. This calculation can be received at the decision-making and execution level. and Alternatively, preliminary calculations can be performed at the T-type smart connector, and the residual value can be reported. Finally, based on the absolute value and sign of the current residual, i.e., if... Determine the health of the line; if This indicates an abnormally low current; if The system determined that there was an abnormally high current (possibly due to insulation damage, partial short circuit, or leakage). The current residual judgment threshold is used to classify and judge the health status of the line and identify potential fault types. The decision execution layer determines the appropriate threshold based on the preset threshold. Logical judgments are made, and the threshold can be dynamically adjusted or adaptively learned based on factors such as system operating environment, cable type, and load characteristics.

[0073] This application's solution utilizes the standard resistance value fixed in the standardized prefabricated quick-connect cables in the physical layer and the synchronous voltage data collected by each voltage monitoring terminal in the sensing layer to accurately calculate the measured and estimated current of each cable segment. By comparing this estimated current with the theoretical current obtained from the node data acquisition layer, the current residual is obtained. This method tightly integrates the physical characteristics of the line with real-time operational data, enabling the decision-making execution layer to accurately determine the health status of the line based on quantified current deviations. When the current residual is within a preset threshold... When the residual is within the specified range, the line is considered healthy. When the residual exceeds the threshold, the system can further distinguish between abnormal current (such as high impedance faults) and abnormal current (such as insulation damage, partial short circuits, or leakage), thus providing a clear basis for subsequent fault location and handling. This current residual analysis based on voltage difference reconstruction effectively compensates for the shortcomings of relying solely on total current sampling or simple voltage monitoring, improving the precision and accuracy of fault diagnosis.

[0074] Preferably, the node current residual analysis includes: defining the first... The inflow current at each T-node is The outflow current is The load current is ; Calculate the node current residual ;like If the leakage current exceeds the preset leakage current threshold, it is determined that a leakage current or grounding fault has occurred at the node or its downstream.

[0075] In the above node current residual analysis, it is first necessary to define the first... Current flowing into each T-node Outflow current and load current A T-junction, in a tree or daisy-chain topology, refers to a physical connection point where cables branch off or connect to loads. Inflow current. This refers to the total current flowing into the T-node from the upstream segment. It can be calculated by reconstructing the voltage difference of the upstream segment or measured by a current sensor installed at the appropriate location. Outflow current. This refers to the total current flowing from the T-node to all its downstream segments. It can be calculated by reconstructing the voltage difference across all downstream segments, or measured by current sensors installed at appropriate locations. Load current. This refers to the current consumed by the load devices directly connected to the T-node. This current is typically measured by the load devices' built-in current sampling circuit and reported to the main control unit. After obtaining the above current value, the node current residual is calculated. Node current residual This is the difference between the current flowing into the T-node and the sum of the current flowing out of the node and its direct loads. According to Kirchhoff's Current Law, at a healthy node, the inflow current should equal the sum of the outflow current and the load current; therefore, the node current residual is... Under normal circumstances, it should be close to zero.

[0076] Subsequently, the calculated nodal current residuals are... The current difference is compared to a preset leakage current threshold. The preset leakage current threshold is a pre-defined upper limit of the current difference used to determine whether a node has a leakage current or a ground fault. This threshold can be set according to the system design requirements, the insulation level of the line, and actual operating experience; for example, it can be set to tens to hundreds of milliamps. When the node current residual... When the absolute value exceeds the preset leakage threshold, the system can determine that the T-node or its directly downstream connected equipment has experienced leakage or grounding faults.

[0077] This application's solution effectively compensates for the shortcomings of relying solely on line segment voltage difference reconstruction algorithms for fault diagnosis by introducing node current residual analysis into a two-wire power supply control system. Specifically, in a topology constructed using standardized prefabricated quick-connect cables and T-type smart connectors at the physical layer, the main control unit and voltage monitoring terminal at the sensing layer can collect the voltage values ​​of each node and ensure data accuracy through a time synchronization mechanism. Based on this, the aforementioned voltage difference reconstruction algorithm can calculate the measured and estimated current of each line segment. Node current residual analysis is based on these line segment current data and the load current data reported by the load devices, applying Kirchhoff's current law at each T-type node. By precisely defining and obtaining the current flowing into the T-type node... Outflow T-shaped node and load Based on the current, the system can calculate the current balance of the T-node. When the calculated node current residual... When the value deviates significantly from zero and exceeds the preset leakage threshold, it directly indicates that there is an abnormal current loss at the T-node, thus accurately indicating that there is a leakage or grounding fault in the node itself or its directly connected downstream equipment. This method refines fault detection from the line segment level to the node level, making fault location more accurate and able to distinguish specific types of faults, rather than just abnormal current on the line segment.

[0078] Preferably, the step-by-step investigation strategy includes:

[0079] Main line diagnosis: Calculating the estimated current of the main line based on voltage difference reconfiguration Calculate the residual of node A. If the threshold is exceeded, the main line is considered to be faulty;

[0080] Branch diagnosis: As the input current of node A, calculate the residual between the theoretical current and the measured estimated current of the downstream branch, and determine the health status of the branch.

[0081] Multi-level expansion: Repeated branch diagnostic logic, proceeding downwards level by level until all end nodes are reached;

[0082] in, This is the power supply voltage. For the voltage at node A, The standard resistor for prefabricated cables in the main trunk line. To calculate the current of the main line based on voltage difference reconstruction, This represents the total current measured at the power supply terminal. Let the current residual at node A be... Let be the current flowing into node A.

[0083] The main line diagnosis is the first step in the hierarchical troubleshooting strategy. Its purpose is to initially assess the health of the power supply system's main lines. By conducting a preliminary diagnosis of the main lines, it's possible to quickly determine if a fault occurs on the main lines, thus avoiding unnecessary troubleshooting of numerous branch lines and improving diagnostic efficiency. This diagnostic process includes calculating the main line current based on voltage difference reconstructing. This step utilizes Ohm's law to calculate the estimated current flowing through the main line based on the voltage difference across the two ends and the known standard resistance value of the cable segment. This estimated current, obtained from the physical characteristics of the line and voltage measurements, is a crucial basis for determining the health status of the main line. For example, the main control unit can obtain the voltage from the power supply voltage sensor. Obtain from the voltage monitoring terminal at node A It also queries the system database for the standard resistance of the prefabricated trunk cable. Then, the above calculations are performed; alternatively, a calculation module can be pre-set in the main control unit, which receives real-time voltage data and pre-stored resistance data, and automatically completes the calculation of the estimated current. Subsequently, the residual of node A is calculated. The residual at node A represents the difference between the measured total current at the power supply end and the calculated current of the main line. This residual reflects whether there is abnormal current loss or gain in the main line itself, and is a direct indicator for judging main line faults. For example, the main control unit will collect the total current at the power supply end in real time. With the calculated Perform a subtraction operation to obtain the residual value; alternatively, the decision execution layer can be configured with a comparator to continuously monitor... and The system calculates the difference between the residuals and outputs it as the residual for node A. If this residual exceeds a threshold, a trunk line fault is determined. By comparing the calculated residual for node A with a preset fault threshold, the system can automatically determine whether a trunk line fault has occurred. This avoids errors and delays from manual judgment and improves the accuracy and real-time performance of the diagnosis. For example, a threshold judgment module can be set up inside the decision execution layer. When the absolute value of the residual exceeds the preset fault threshold, a trunk line fault alarm is triggered; or, the threshold can be dynamically adjusted based on historical data, line type, and environmental conditions to adapt to different operating conditions and improve the robustness of the judgment. After the trunk line diagnosis is completed, if the trunk line is determined to be healthy or the fault has been located, the system will enter the branch line diagnosis stage. In this stage, the system will use... As the input current for node A, this step is crucial for achieving a step-by-step investigation "from root to leaf." It uses the current calculated from the upstream trunk line as the input current for the current node, ensuring the continuity and logic of the diagnosis, and enabling the diagnosis of downstream branches to be based on accurate upstream information. For example, when the decision-making execution layer performs branch diagnosis, it uses the current calculated during the trunk line diagnosis stage... The current value is passed to the branch diagnostic module as its initial input current; alternatively, the system maintains a current status table, updating the input current value of each node in real time for use by downstream diagnostic modules. Next, the residual between the theoretical current and the measured estimated current of the downstream branch is calculated to determine the branch's health status. By comparing the branch's theoretical current (e.g., based on the current reported by the load device) with the measured estimated current obtained through voltage difference reconstruction, abnormal current conditions in the branch can be detected, thus determining whether a fault exists in the branch. For example, for each downstream branch, the system obtains its theoretical current value (e.g., measured from the load device's built-in current sampling circuit and reported to the main control unit via the communication network), calculates the measured estimated current using the voltage difference across the branch and a standard resistance, and then calculates the residual between the two; alternatively, the decision execution layer can maintain a state machine for each branch, updating the branch's health status, such as "normal," "abnormally low," or "abnormally high," based on the magnitude and duration of the current residual. To ensure comprehensive diagnosis of the entire power supply network, this strategy also includes a multi-level expansion mechanism. This mechanism repeats the branch diagnosis logic described above, checking downwards level by level until all end nodes are reached. This means that regardless of how many levels of branches the power supply network has, the system can systematically and progressively cover faults. For example, the decision execution layer can use a depth-first search or breadth-first search algorithm to traverse the entire topology and execute diagnosis logic for each node and its connected branches; alternatively, the system can maintain a node queue, taking one node from the queue for diagnosis each time and adding its downstream nodes to the queue until the queue is empty.

[0084] The proposed solution decomposes the fault diagnosis process into clearly defined main line and branch line diagnoses, and employs a root-to-leaf hierarchical investigation approach, significantly improving the efficiency and accuracy of fault location. By first investigating the main line, the fault range can be quickly identified, avoiding unnecessary testing of numerous healthy branches. Subsequently, using upstream estimated current as downstream input ensures the logical coherence and data accuracy of each diagnostic level, thereby enabling precise identification of specific faulty segments or nodes. This layered, progressive diagnostic method effectively solves the problems of low efficiency and insufficient accuracy in fault location within complex power supply networks.

[0085] Preferably, the time synchronization mechanism adopts any of the following methods: the master control unit periodically broadcasts synchronization frames through the communication bus, and each node starts sampling after receiving the synchronization frame after the same preset delay; the system adopts the IEEE 1588 precise time protocol; each node has a built-in real-time clock, which is periodically calibrated by the master control unit.

[0086] Preferably, the T-type intelligent connector terminal adopts a power supply method that combines line self-powering and backup energy storage: the internally integrated wide-input DC-DC conversion circuit draws power from the line as the main power source, while a supercapacitor or rechargeable battery is connected in parallel as a backup power source; when the line voltage drops or is broken, the backup power source is automatically put into operation to maintain the node operation for at least 30 seconds and realize the reporting of the fault at the end of the fault period.

[0087] Preferably, it also includes an adaptive monitoring strategy:

[0088] When the total current at the power supply terminal exceeds the preset threshold, the voltage drop residual analysis mode is entered.

[0089] When the total current at the power supply end is less than the preset threshold, switch to the level consistency monitoring mode;

[0090] In no-load mode, the system control power supply terminal actively injects constant current test pulses, uses the voltage difference during the pulse to calculate the actual line resistance, and compares it with the standard resistance to determine the line health.

[0091] The specific steps for determining the health of a circuit by comparing it with a standard resistor are as follows:

[0092] Low current / no-load mode (level consistency monitoring mode)

[0093] Triggering condition: Actual measured total current at the power supply terminal (Adaptive operating condition switching current threshold, typical value 1A), including two cases: low current load and complete no-load.

[0094] Execution process:

[0095] Low current mode ( )

[0096] The system can still try to use the voltage difference reconstruction algorithm, but due to the small current and small voltage drop, the measurement error may be large.

[0097] At this point, level consistency monitoring is used as an auxiliary method:

[0098] Compare the voltages of adjacent nodes and The difference, if ( If the voltage difference threshold (typically 0.5V) is set, then the cable segment is determined to have abnormal leakage or poor insulation (because under no load or light load, the voltage drop of a healthy cable should be close to 0). If the cable is basically normal, a minor high-resistance fault cannot be ruled out (further diagnosis requires active pulse injection).

[0099] No-load mode (actual measured total current at the power supply end) )

[0100] With all loads off, there is almost no current in the cables;

[0101] Level consistency monitoring: If the voltage difference between the two ends of a certain cable segment... The system was found to have an abnormal leakage current (because the voltage at both ends of a healthy cable should be equal when it is unloaded).

[0102] Open circuit diagnosis: If And the voltage at a downstream node Significantly lower than the voltage at the upstream node (For example If the cable segment is open, then the cable segment is considered to be broken.

[0103] Active signal injection (for accurate diagnosis under no-load conditions)

[0104] When the system is in idle mode for an extended period (such as during the night or early morning), the system proactively performs the following steps to improve diagnostic capabilities:

[0105] Step 1: Inject constant current test pulse

[0106] The power supply main control unit controls the output stage to inject a brief constant current pulse into the line, with the following parameters:

[0107] Current amplitude (Adjustable according to cable specifications)

[0108] Pulse width (Short enough to avoid overheating, long enough to stabilize sampling);

[0109] Step 2: Measure the response voltage difference

[0110] Each T-type node synchronously samples the voltage during the pulse.

[0111] For the cable segment under test (e.g., segment AB), calculate the voltage difference during the pulse period. ;

[0112] Step 3: Calculate the actual line resistance ;

[0113] Step 4: Compare with a standard resistor

[0114] The standard resistance of this cable segment is known from the manufacturer. (Already embedded in the electronic tag);

[0115] Calculate the rate of change of resistance: ;

[0116] Step 5: Health Assessment

[0117] like ( The cable health is determined by a preset threshold (typically 5%).

[0118] like If the cable resistance is determined to be increased due to aging, corrosion, or oxidation of the connectors, a preventative maintenance warning will be issued.

[0119] like If the actual resistance is less than the nominal value, it is determined that there is a partial short circuit or insulation damage, and a fault alarm is issued.

[0120] Step 6: Return to normal

[0121] After the pulse ends, the system returns to the no-load monitoring mode and waits for the next active diagnostic cycle (e.g., once per hour).

[0122] Preferably, it also includes an anti-theft and unauthorized access monitoring module; the module determines whether there is unauthorized load access based on voltage difference reconstruction and node current residual analysis according to the following steps:

[0123] Step 1: Calculate the outlier confidence levels for the following four factors: Factor A: Node power imbalance confidence level Factor B: Impedance characteristic matching confidence level Factor C: Confidence level of historical current baseline deviation Factor E: Confidence level of timing correlation of load switch events ;

[0124] Step 2: Calculate the overall confidence level using a multi-factor fusion formula: ,in, The overall confidence level is set and the value range is [0,1]. As a factor index, and taking A, B, C, E, For the first The preset weights of each factor and satisfy the following conditions: ; For the first The outlier confidence level of each factor, with a value range of [0,1];

[0125] Step 3: Assemble the overall confidence level With the first preset threshold and the second preset threshold Comparison: If If it determines that there is electricity theft or illegal access, an alarm will be triggered; if If marked as suspected, an active pulse detection will be automatically triggered for further confirmation; if The result is normal.

[0126] Among them, the confidence level of node power imbalance Calculated using the following formula:

[0127]

[0128] in, Let be the confidence level for node power imbalance, with a value range of [0,1]. For nodes The power residual, For low power threshold, High power threshold;

[0129] like Duration ,but Upward correction, of which, For duration, For duration threshold, This is the time scaling factor (reference time).

[0130] Impedance characteristic matching confidence Calculated using the following formula:

[0131]

[0132] in, The impedance characteristic matching confidence level has a value range of [0,1]. branch road The relative impedance deviation, For low threshold impedance deviation, For high impedance deviation threshold, For real-time impedance, This is the reference impedance for this branch under healthy conditions.

[0133] Historical current baseline deviation confidence level ; Calculated using the following formula: ,in, The historical current baseline deviation confidence level is defined, with values ​​ranging from [0,1]. The standard normal cumulative distribution function is... The absolute value of the relative deviation of the current and , This represents the mean of the historical baseline deviation. This is the historical average inflow current at the same time over the past few days. It is a small positive number (typical value is 0.01A). The standard deviation of the historical baseline;

[0134] Confidence of event timing correlation of load switch Calculated using the following formula: ,in, The confidence level for the timing correlation of load switch events is defined, and its value ranges from [0,1]. This represents the absolute value of the current step amplitude at the moment without instruction. The step current threshold, The time difference between the step occurrence time and the most recent authorized instruction. This is the time decay constant (in seconds).

[0135] If no step jump occurs again in several consecutive cycles, then Decline by exponential rate: ,in, For time indexing, The time interval between two tests, This is the decay time constant (in seconds).

[0136] Among them, the confidence level of node power imbalance This reflects the deviation between the actual power and the theoretical power at a node. When an illegal load is connected, the actual power consumption of the node will exceed expectations, leading to power imbalance. The calculation method can be by comparing the node's real-time power with the theoretical power calculated based on known loads and line conditions, or by comparing the estimated power flowing into the node with the sum of the power consumed by the outflowing node and known loads. Impedance characteristic matching confidence level. This measures the degree of matching between a real-time measurement of line or node impedance and a baseline value under healthy conditions. Unauthorized load connections typically alter the equivalent impedance of a local line, causing a deviation in impedance characteristics. This can be calculated in real-time by injecting a test signal and measuring the response, then comparing it to a pre-stored normal impedance model. Historical current baseline deviation confidence level. The deviation of the current value from the historical normal current pattern was assessed. Electricity theft or illegal access often manifests as an abnormal increase or fluctuation in current, inconsistent with the historical baseline of normal system operation. The calculation method can be to establish a current baseline model for different time periods (e.g., hourly, daily, weekly) through statistical analysis of historical data, and then compare the real-time current with this model. The confidence level of the timing correlation of load switch events was also assessed. This method is used to detect step changes in current that occur without system commands. Normal load switching operations are typically commanded by the system, while the switching behavior of illegally connected loads is unrelated to system commands, manifesting as sudden current changes without instructions. The calculation method involves monitoring the instantaneous rate of change of current to identify step events and verifying the existence of corresponding legitimate switching commands in the system log. A multi-factor fusion formula is used to calculate the overall confidence level. This step involves weighted fusion of the five anomaly confidence levels mentioned above to obtain a more comprehensive and reliable overall judgment index. The fusion formula uses a product form, which can effectively amplify multiple weak anomaly signals while avoiding misjudgment based on a single factor. The weights are... The sensitivity and reliability of unauthorized access can be preset based on various factors, for example, through optimization using expert experience or machine learning methods. The overall confidence level will then be considered. With the first preset threshold and the second preset threshold The comparison is then performed. This step is the final decision-making stage. By comparing the calculated overall confidence level with two preset thresholds, anomalies are categorized into three levels: "definitely anomaly," "suspected anomaly," and "normal." This tiered decision-making mechanism can improve the accuracy of diagnosis and provide opportunities for further confirmation of suspected cases. Specifically, if... If it determines that there is electricity theft or illegal access, an alarm will be triggered; if If marked as suspected, an active pulse detection will be automatically triggered for further confirmation; if The result is normal.

[0137] The anti-theft and unauthorized access monitoring module of this application works by integrating the existing voltage difference reconstruction algorithm and node current residual analysis capabilities of a two-wire power supply control system, and introducing multi-dimensional anomaly feature analysis to accurately identify unauthorized load access. Specifically, the module first uses voltage and total power supply current data collected by the system perception layer, combined with theoretical values ​​provided by the node data acquisition layer, to calculate the actual current of each cable segment using the voltage difference reconstruction algorithm, and evaluates the current balance of each node through node current residual analysis. Based on this, the module further calculates five anomaly confidence levels: node power imbalance confidence level, impedance characteristic matching confidence level, historical current baseline deviation confidence level, active pulse reflection characteristic confidence level, and load switching event timing correlation confidence level. These confidence levels quantify the degree of deviation between the system's operating state and the normal mode from different perspectives. For example, unauthorized access can lead to abnormal increases in node power, changes in line impedance, current deviation from the historical baseline, abnormal reflection of test pulses, and current steps without instructions. Subsequently, these independent confidence levels are calculated into a comprehensive confidence level using a multi-factor fusion formula. This fusion formula, through weighted multiplication, effectively aggregates information from different detection dimensions, enhancing sensitivity to weak anomaly signals while reducing the false alarm rate that might arise from a single detection method. Finally, this comprehensive confidence level is compared with preset high and low thresholds, clearly classifying the system status into three categories: "electricity theft or illegal access," "suspected anomaly," and "normal." For suspected cases, the system automatically triggers more in-depth active pulse detection for secondary confirmation, ensuring diagnostic accuracy while avoiding unnecessary alarms. This multi-dimensional, hierarchical decision-making mechanism enables the system to effectively identify and address electricity theft and illegal access issues that are difficult to cover with traditional fault diagnosis, significantly improving the safety and reliability of the power supply system.

[0138] Preferably, the anti-electricity theft and illegal access monitoring module further includes a confidence correction layer. This layer dynamically corrects the overall confidence level based on one or more of the following environmental and historical behavioral factors to obtain a final confidence level. The final confidence level is then used to replace the overall confidence level for threshold comparison: Factor F1: Soil moisture influence factor. ,in, Soil moisture influencing factors To preset the sensitivity coefficient, This refers to the measured or estimated volumetric water content of the soil. For reference humidity, To take the positive part of the function, that is, when the part inside the parentheses is negative, it takes the value 0;

[0139] Factor F2: Circuit thermal aging accumulation factor: , ,in, This is the cumulative factor for thermal aging of the circuit. The aging index, To calculate the total power-on time, For cable design life, The temperature acceleration factor (unit: per degree Celsius (1 / ℃)). The average operating temperature, For reference temperature, This is the aging inhibition coefficient;

[0140] Factor F3: Diurnal / Seasonal Pattern Factor ,in, For day / night / seasonal pattern factors, The preset suppression coefficient, This is an indicator function that takes the value 1 if any of the following conditions are met, and 0 otherwise:

[0141] The current time is late at night (0:00 to 5:00).

[0142] The current time is within the historically low load period for the same period (i.e.) , (Take 5% of the system's rated current).

[0143] Factor F4: Consistency factor in the behavior of neighboring nodes. ,in, This serves as a consistency factor for the behavior of adjacent nodes. For the preset enhancement coefficient, This represents the standard deviation of the anomaly confidence scores of each child node under the same parent node. The standard deviation of the confidence level of historical anomalies across the entire network;

[0144] Factor F5: Load switch fatigue index: , ,in, For load switch fatigue factor, The fatigue index. This represents the average number of times the device is switched on and off per day. To design the rated number of switching cycles, This is the fatigue inhibition coefficient;

[0145] Final confidence level correction formula: ,in, The final confidence level is defined with a value range of [0,1]. To assess the overall confidence level, For factor index, No. One correction factor, For the first The enable flag for each factor is either 0 or 1, where 0 indicates that the factor does not participate in the correction and 1 indicates that it does.

[0146] by With the first preset threshold and the second preset threshold Comparative decision-making;

[0147] For each factor ( =1, 2, 3, 4, 5 (corresponding to F1~F5 respectively), their activation flags. Determined according to the following rules:

[0148] Rule 1: If the system has the data source required for the factor (for example, factor F1 requires soil moisture data, factor F2 requires temperature sensor data, factor F3 requires historical current data, factor F4 requires data from multiple child nodes, and factor F5 requires records of the number of times a switch is turned on), then The initial value can be set to 1, otherwise it should be set to 0.

[0149] Rule 2: System administrators can manually configure the activation status of each factor through the host computer software, overriding the default value;

[0150] Rule 3: The system can automatically adjust based on historical false alarm rates: If enabling a certain factor leads to a decrease in the overall confidence level... If the consistency with the manual review results is below a threshold (e.g., 80%), the factor will be automatically disabled and logged.

[0151] Rule 4: For factors lacking real-time data (e.g., factor D requires active pulse testing, but cannot be injected during busy periods), temporarily set... =0, will be restored after the condition is met.

[0152] Among them, soil moisture influencing factors This method is used to quantify the impact of soil moisture on the electrical characteristics of power lines. Soil moisture is an important environmental factor affecting the insulation performance and grounding resistance of underground cables. High humidity may lead to a decrease in insulation resistance or an increase in leakage current, thus affecting the results of voltage difference reconstruction and current residual analysis. This factor is determined by measured or estimated soil volumetric water content. Compared with reference humidity The correction factor is calculated based on the difference in humidity. When the humidity is higher than the reference value, the correction factor will be adjusted accordingly to avoid misjudgments caused by changes in environmental humidity. For example, real-time data can be obtained by deploying soil moisture sensors in key areas, or estimation can be made by combining meteorological forecast data and historical experience models. Line thermal aging cumulative factor. This factor is used to assess the degree of aging of cables due to long-term operation and temperature stress. Cable aging causes changes in parameters such as resistivity and insulation performance, which in turn affects the accuracy of fault diagnosis. This factor is assessed through an aging index. To reflect the cumulative aging effect of cables, the aging index is used. The cumulative power-on time was taken into account. Cable design life and average operating temperature Factors such as the cumulative energization time of each cable segment and the acquisition of operating temperature data via built-in or external temperature sensors, combined with a preset temperature acceleration coefficient, are considered. To calculate the aging index. Diurnal / seasonal pattern factor. This factor is used to reflect the typical variation of system load in different time periods or seasons. During certain specific periods (such as late at night) or seasons (such as winter off-peak periods), the system load is usually lower. If abnormal current fluctuations occur during these times, the assessment of the degree of abnormality should differ from that during peak periods. This factor is indicated by a function... This allows the system to identify whether it is currently in a low-load period, thereby appropriately suppressing or adjusting the confidence level. For example, the system can have a built-in schedule to define late-night periods, or dynamically identify low-load periods by analyzing historical current data. Neighbor node behavior consistency factor. This factor is used to assess the correlation between local and global anomalies. If the anomaly confidence of a node is significantly higher than that of its neighboring nodes or the average level of the entire network, the anomaly is more likely to be localized. Conversely, if multiple neighboring nodes exhibit anomalies simultaneously, it may indicate a broader problem. This factor is calculated by comparing the standard deviation of the anomaly confidence of each child node under the same parent node. Standard deviation of the confidence level of historical anomalies across the entire network To provide corrections. For example, the system can maintain a network topology and calculate the anomaly confidence level of each node in real time, thereby calculating the local and global standard deviations. Load switch fatigue index. This factor is used to account for the impact of frequent load switching operations on system diagnostics. Frequent switching actions can generate transient current or voltage fluctuations, which may be misinterpreted as abnormalities in some cases. This factor is addressed through the fatigue index. To quantify the activity level of load switches, including the fatigue index. Based on average daily number of switching times With respect to the design rated number of switching cycles The comparison. For example, the system can record the number of times each load device is switched on and off and calculate its daily average. Final confidence correction formula. The initial overall confidence level With the above correction factors The final confidence level is obtained by fusion. .in, The enable flag for each factor allows the system to selectively enable or disable specific correction factors based on actual conditions. For example, if the data source required by a factor is unavailable, or if the factor is deemed unsuitable in a specific scenario, its enable flag can be set to 0, preventing it from participating in correction. Factor Enable Flags The determination of factors follows a series of rules to ensure the flexibility and accuracy of the correction process. Rule 1 ensures that factors can only be enabled when the system has the required data source. Rule 2 allows system administrators to manually configure the enabling status of factors based on experience or specific needs. Rule 3 introduces an automatic adjustment mechanism based on historical false alarm rates, enabling the system to continuously optimize correction strategies through learning and improve diagnostic accuracy. Rule 4 handles temporary disabling due to missing real-time data, ensuring the system's adaptability in dynamic environments.

[0153] This application's solution introduces a confidence correction layer, enabling the anti-theft and illegal access monitoring module to dynamically and intelligently adjust the overall confidence level based on real-time environmental conditions and system operating status. When the anti-theft and illegal access monitoring module initially determines an anomaly, the confidence correction layer refines the anomaly confidence level based on various factors such as current soil moisture, line aging, load patterns during the current time period, consistency of behavior among adjacent nodes, and the activity level of load switches. For example, if a slight leakage is detected in a humid environment, the correction layer may appropriately lower its anomaly confidence level to avoid false alarms; if an abnormally large current is detected during an abnormal load period, the correction layer may increase its anomaly confidence level, prompting the system to intervene earlier. This dynamic correction mechanism, combined with the multi-factor fusion judgment of the anti-theft and illegal access monitoring module, forms a more intelligent and robust diagnostic system.

[0154] A fault diagnosis method for a two-wire power supply control system is implemented using the aforementioned system.

[0155] It should be noted that, in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus.

[0156] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A fault diagnosis system for a two-wire power supply control system, characterized in that, include: Physical layer: Standardized prefabricated quick-connect cables and T-type smart connectors are used to construct a tree-like or daisy-chain topology; each prefabricated cable segment has its length, standard resistance value and unique electronic tag fixed at the factory; The sensing layer includes a main control unit installed at the power supply end and voltage monitoring terminals distributed at each T-type smart connector. The main control unit has a total current sampling function, and each voltage monitoring terminal collects the voltage value of the corresponding node. All monitoring terminals have a time synchronization mechanism. In the sensing layer, only the power supply end is equipped with a current sensor. The load current is measured by the current sampling circuit built into the load device and reported to the main control unit through the communication network. No current sensor is installed at the T-type node. Node data acquisition layer: Real-time acquisition of theoretical voltage and theoretical current values ​​for each node; Decision execution layer: Fault diagnosis and location are performed using voltage difference reconstruction algorithm and node current residual analysis. The decision execution layer adopts a step-by-step investigation strategy from root to leaf, taking the estimated current obtained by voltage difference reconstruction of upstream nodes as the known input current of downstream nodes, and calculating the residual between the theoretical current and the measured estimated current of each branch step by step to achieve accurate fault location.

2. The fault diagnosis system for a two-wire power supply control system according to claim 1, characterized in that, The voltage difference reconstruction algorithm includes: obtaining the voltage between two adjacent nodes. , ; Calculate the voltage difference According to the standard resistance of this prefabricated cable section Calculate the measured and estimated current Obtain the theoretical current of this cable segment. ; Calculate current residual ;like Determine the health of the line; if This indicates an abnormally low current; if The system determined that there was an abnormally high current; among them, The threshold for judging current residual is set.

3. The fault diagnosis system for a two-wire power supply control system according to claim 1, characterized in that, The node current residual analysis includes: defining the first... The inflow current at each T-node is The outflow current is The load current is ; Calculate the node current residual ;like If the leakage current exceeds the preset leakage current threshold, it is determined that a leakage current or grounding fault has occurred at the node or its downstream.

4. The fault diagnosis system for a two-wire power supply control system according to claim 1, characterized in that, It also includes an adaptive monitoring strategy: when the total current at the power supply end is greater than a preset threshold, it enters the voltage drop residual analysis mode; when the total current at the power supply end is less than a preset threshold, it switches to the level consistency monitoring mode. In no-load mode, the system control power supply terminal actively injects constant current test pulses, uses the voltage difference during the pulse to calculate the actual line resistance, and compares it with the standard resistance to determine the line health.

5. The fault diagnosis system for a two-wire power supply control system according to claim 1, characterized in that, The step-by-step investigation strategy includes: Main line diagnosis: Calculating the estimated current of the main line based on voltage difference reconfiguration Calculate the residual of node A. If the threshold is exceeded, the main line is considered to be faulty; Branch diagnosis: As the input current of node A, calculate the residual between the theoretical current and the measured estimated current of the downstream branch, and determine the health status of the branch. Multi-level expansion: Repeated branch diagnostic logic, proceeding downwards level by level until all end nodes are reached; in, This is the power supply voltage. For the voltage at node A, The standard resistor for prefabricated cables in the main trunk line. To calculate the current of the main line based on voltage difference reconstruction, This represents the total current measured at the power supply terminal. Let the current residual at node A be... Let be the current flowing into node A.

6. The fault diagnosis system for a two-wire power supply control system according to claim 1, characterized in that, It also includes a module for monitoring electricity theft and unauthorized access; this module, based on voltage difference reconstruction and node current residual analysis, determines whether there is unauthorized load access according to the following steps: Step 1: Calculate the outlier confidence levels for the following four factors: Factor A: Node power imbalance confidence level Factor B: Impedance characteristic matching confidence level ; Factor C: Confidence level of historical current baseline deviation ; Factor E: Confidence level of timing correlation of load switching events ; Step 2: Calculate the overall confidence level using a multi-factor fusion formula: ,in, The overall confidence level is set and the value range is [0,1]. As a factor index, and taking A, B, C, E, For the first The preset weights of each factor and satisfy the following conditions: ; For the first The outlier confidence level of each factor, with a value range of [0,1]; Step 3: Assemble the overall confidence level With the first preset threshold and the second preset threshold Comparison: If If it determines that there is electricity theft or illegal access, an alarm will be triggered; if If marked as suspected, an active pulse detection will be automatically triggered for further confirmation; if The result is normal.

7. The fault diagnosis system for a two-wire power supply control system according to claim 6, characterized in that, The anti-electricity theft and unauthorized access monitoring module also includes a confidence level correction layer. This layer dynamically corrects the overall confidence level based on one or more environmental and historical behavioral factors to obtain a final confidence level, which is then used to replace the overall confidence level for threshold comparison. Factor F1: Factors affecting soil moisture: ,in, Soil moisture influencing factors To preset the sensitivity coefficient, This refers to the measured or estimated volumetric water content of the soil. For reference humidity, To take the positive part of the function, that is, when the part inside the parentheses is negative, it takes the value 0; Factor F2: Circuit thermal aging accumulation factor: , ,in, This is the cumulative factor for thermal aging of the circuit. The aging index, To calculate the total power-on time, For cable design life, The temperature acceleration coefficient, The average operating temperature, For reference temperature, This is the aging inhibition coefficient; Factor F3: Day / Night / Seasonal Pattern Factor: ,in, For day / night / seasonal pattern factors, The preset suppression coefficient, This is an indicator function that takes the value 1 if any of the following conditions are met, and 0 otherwise: the current time is during the late night period; The current time is within a historically low load period for this time of year; Factor F4: Consistency factor in the behavior of neighboring nodes. ,in, This serves as a consistency factor for the behavior of adjacent nodes. For the preset enhancement coefficient, This represents the standard deviation of the anomaly confidence scores of each child node under the same parent node. The standard deviation of the confidence level of historical anomalies across the entire network; Factor F5: Load switch fatigue index: ,, ,in, For load switch fatigue factor, The fatigue index. This represents the average number of times the device is switched on and off per day. To design the rated number of switching cycles, This is the fatigue inhibition coefficient; Final confidence level correction formula: ,in, The final confidence level is defined with a value range of [0,1]. To assess the overall confidence level, For factor index, No. One correction factor, For the first The enable flag for each factor is either 0 or 1, where 0 indicates that the factor does not participate in the correction and 1 indicates that it does. by With the first preset threshold and the second preset threshold Make a comparative decision.

8. The fault diagnosis system for a two-wire power supply control system according to claim 7, characterized in that, For each factor , =1, 2, 3, 4, 5 correspond to F1~F5 respectively, and their activation flags are... Determined according to the following rules: Rule 1: If the system has the data source required for the factor, then The initial value can be set to 1, otherwise it should be set to 0. Rule 2: System administrators can manually configure the activation status of each factor through the host computer software, overriding the default value; Rule 3: The system can automatically adjust based on historical false alarm rates: If enabling a certain factor leads to a decrease in the overall confidence level... If the consistency with the manual review result is below the threshold, the factor will be automatically disabled and logged. Rule 4: For factors lacking real-time data, temporarily set... =0, will be restored after the condition is met.

Images