Internet of Things-based Rail Transit Electromechanical Equipment Control System

Abstract

This invention specifically relates to an IoT-based control system for rail transit electromechanical equipment, encompassing the field of rail transit electromechanical equipment control technology. It includes: a topology self-discovery and map building module; a routine monitoring and performance learning module; and a fault diagnosis and reconfiguration triggering module. In this invention, a fault coefficient determination mechanism enables accurate identification of fault signals, effectively solving the problem of false alarms in traditional systems. The system integrates outlier values ​​from operating condition matching with the asynchronous rate of strongly correlated nodes, quantifying the fault coefficient using the area of ​​a triangle. It verifies signal authenticity from both static operating condition adaptation and dynamic group collaboration dimensions, accurately distinguishing between sensor-specific faults and actual equipment faults, thus avoiding unnecessary downtime or reconfiguration actions.

Description

Technical Field

[0001] This invention relates to the field of rail transit electromechanical equipment control technology, and in particular to a rail transit electromechanical equipment control system based on the Internet of Things. Background Technology

[0002] Rail transit electromechanical equipment (such as drainage pumps, solenoid valves, and various sensors) is the core infrastructure for ensuring the safe operation of rail lines. Its stable operation is directly related to the continuity of key functions such as flood control and drainage, equipment cooling, and power supply, and plays an irreplaceable role in the safe operation of rail transit.

[0003] The existing rail transit electromechanical equipment control systems mostly adopt a fixed logic automatic control mode. The topology relationship needs to be manually configured after on-site survey. When equipment is replaced, pipeline is modified, or new equipment is added, the topology structure changes and reprogramming and debugging are required. This not only results in poor adaptability and low maintenance efficiency, but also makes the control logic prone to failure due to human error.

[0004] Meanwhile, the performance baseline of traditional systems is mostly a globally unified standard, without being subdivided according to actual operating conditions such as operating frequency, liquid level, and flow requirements. This makes it easy for false alarms or missed alarms to occur when operating conditions fluctuate, affecting the stability of system operation.

[0005] Traditional control systems have significant limitations in their fault diagnosis mechanisms. They often rely on single sensor signals for judgment, failing to consider the interrelationships and coordination between devices. Sensor drift, damage, or electromagnetic interference can easily lead to misjudgments, resulting in unnecessary equipment downtime or reconfiguration. Furthermore, after a fault occurs, the system often only issues an alarm signal awaiting manual intervention, lacking proactive fault tolerance and posing a high risk of core function interruption. Even if some systems have backup equipment activation capabilities, they lack scientific alternative path optimization algorithms and safe execution sequences. This can easily lead to secondary faults due to unreasonable path selection or improper operation sequence, further expanding the scope of the fault's impact and failing to meet the high reliability and low interruption rate requirements of rail transit electromechanical equipment.

[0006] Therefore, an IoT-based control system for rail transit electromechanical equipment is proposed to address the aforementioned problems. Summary of the Invention

[0007] The purpose of this invention is to propose an Internet of Things-based control system for rail transit electromechanical equipment in order to solve the above-mentioned problems.

[0008] To achieve the above objectives, the present invention adopts the following technical solution:

[0009] The Internet of Things-based control system for rail transit electromechanical equipment includes:

[0010] The topology self-discovery and map building module is configured to collect sensor data through standardized detection sequences, analyze device correlations through multi-algorithm processing, and generate a highly reliable structured topology model.

[0011] The routine monitoring and performance learning module is configured to optimize control strategies based on topology graphs, continuously collect equipment operation data, and establish performance baselines.

[0012] The fault diagnosis and reconfiguration trigger module is configured to trigger fault signal diagnosis to obtain fault coefficients when the controller receives a device fault feedback signal, or when the comparison results of various real-time data with the performance baseline under the corresponding operating conditions meet the abnormal conditions. Based on the fault coefficients, it determines whether the corresponding fault signal is erroneous. If the fault signal is erroneous, it determines the fault type and analyzes the scope of impact, and triggers the reconfiguration process.

[0013] Preferably, the topology self-discovery and map building module specifically includes:

[0014] After the system is powered on for the first time, it completes the initialization of the operating system and drivers, and automatically triggers topology discovery.

[0015] The central controller first sends instructions to put all actuators into their initial state and calibrate all sensors to eliminate zero drift;

[0016] The collected sensor data is filtered to remove spike noise caused by electromagnetic interference; at the same time, outliers that exceed the sensor's measurement range are also removed.

[0017] Calculate the Pearson correlation coefficient between actuator action and sensor value changes, and determine the correlation strength based on the Pearson correlation coefficient;

[0018] Set a minimum threshold for sensor value changes; only when the change exceeds the threshold and matches the timing of actuator actions is it determined to be associated.

[0019] The topology graph data model structure includes:

[0020] Node table: contains node ID, device type, installation location, device parameters, and health status;

[0021] Edge table: contains edge ID, starting node ID, ending node ID, connection type, association strength, fluid flow direction, and path length.

[0022] Preferably, the routine monitoring and performance learning module specifically includes:

[0023] Based on the topology diagram, devices with the same function are started in rotation according to their runtime; at the same time, the cumulative runtime of each device is recorded.

[0024] By analyzing the flow distribution of each branch through the topology diagram, when the flow of a branch is greater than or equal to a preset percentage of the rated flow, the backup solenoid valve in the same area is opened to divert the flow to the adjacent branch.

[0025] Based on real-time flow and pressure data, the frequency of the water pump and the on / off status of the solenoid valve are dynamically adjusted to ensure that the system operates in the optimal range where the flow meets the demand and energy consumption is minimized.

[0026] Preferably, the process further includes performance baseline learning:

[0027] The data collection dimensions and frequency are as follows:

[0028] Water pump: starting current, steady-state operating current, operating frequency, vibration acceleration, inlet and outlet pressure, operating noise;

[0029] Solenoid valve: opening time, closing time, coil current, pressure difference across the valve;

[0030] Sensor: Measured value, measurement error, response time;

[0031] The baseline establishment method is as follows:

[0032] Collect a preset number of valid data points under normal operating conditions, and ensure that the data covers different operating loads;

[0033] Statistical methods are used to generate baseline ranges. For continuous data, the preset values ​​of the samples are taken as the normal range; for time series data, the average curve of the samples is taken as the baseline curve.

[0034] Baselines are established according to equipment type and operating conditions.

[0035] Preferably, the fault diagnosis and reconfiguration triggering module specifically includes:

[0036] The controller directly receives fault feedback signals from the equipment, including overcurrent protection signals, short circuit alarms, mechanical jamming signals, and power failure signals.

[0037] Real-time data is compared with the performance baseline under the corresponding operating conditions, and is judged to be abnormal if it meets the following conditions:

[0038] The water pump operating current exceeds the baseline range for a preset maximum duration.

[0039] The solenoid valve opening time is extended by a preset percentage compared to the baseline.

[0040] The sensor measurement value fluctuates by a preset percentage beyond the baseline.

[0041] Preferably, the process of obtaining the fault coefficient includes:

[0042] Based on sensors, the operating frequency of the water pump, the suction side liquid level, and the regional flow demand corresponding to the fault signal are obtained, and the sensor baseline range for the corresponding operating condition is determined.

[0043] Operating condition deviation rate = (Target sensor measured value - Baseline median under corresponding operating condition) ÷ Baseline median under corresponding operating condition × 100%;

[0044] Baseline median = (maximum baseline value under the corresponding working condition + minimum baseline value under the corresponding working condition) ÷ 2;

[0045] A preset operating condition deviation rate threshold is set. If the operating condition deviation rate is greater than the operating condition deviation rate threshold, the absolute value of the difference between the two is calculated and recorded as an abnormal value of the operating condition deviation.

[0046] The parameters with the largest changes among the pump operating frequency, suction side liquid level, and regional flow demand are selected as reference conditions. The data changes of the corresponding sensors within a preset time period and the corresponding changes of the operating parameters are obtained for the reference conditions.

[0047] The operating condition response value is obtained by dividing the change in sensor data within a preset time period by the change in operating parameters.

[0048] Obtain the corresponding baseline operating condition response value range, match the operating condition response value with the operating condition response value range, mark the operating condition response value that is not within the operating condition response value range, calculate the difference between it and the operating condition response value range, take the absolute value, and record it as an abnormal operating condition response value.

[0049] After normalizing the abnormal values ​​of operating condition deviation and operating condition response, their values ​​are set to the range of [0,1]. Weighted summation is then performed to obtain the abnormal values ​​of operating condition matching.

[0050] Preferably, the method further includes:

[0051] Starting with the faulty water pump corresponding to the sensor fault signal, traverse the topology graph and filter out all sensor nodes that are strongly associated with the water pump.

[0052] Acquire the node signal values ​​monitored by each sensor node strongly associated with the water pump, and preset the allowable fluctuation range of the node signal values;

[0053] If the allowable fluctuation range of the preset node signal value covers the sensor fault signal, then the sensor corresponding to the allowable fluctuation range of the preset node signal value is marked as a synchronization abnormal sensor.

[0054] After obtaining all the synchronous abnormal sensors, divide by the number of sensor nodes strongly associated with the water pump to get the synchronization rate, and subtract the synchronization rate from 1 to get the asynchronous rate.

[0055] Preferably, after normalizing the abnormal value of the working condition matching and the asynchronous rate, the value range is [0,1]. The abnormal value of the working condition matching and the asynchronous rate are respectively used as the base and the height perpendicular to the base of the triangle to construct the triangle, and the area of ​​the triangle is calculated and recorded as the fault coefficient.

[0056] A preset fault coefficient threshold is set. If the fault coefficient is greater than the fault coefficient threshold, the sensor corresponding to the fault coefficient is determined to be faulty, and the fault signal is incorrect.

[0057] Preferably, the method further includes an impact range analysis:

[0058] Starting with the faulty device, traverse all nodes in the topology graph that are strongly associated with the device to determine the direct impact range; then traverse the weakly associated nodes to determine the indirect impact range.

[0059] The analysis dimensions include the device level, the region level, and the function level;

[0060] Based on the impact level, it includes minor impact, moderate impact, and severe impact;

[0061] The analysis results will be output as an impact range analysis report.

[0062] Preferably, the method further includes:

[0063] The dynamic control reconfiguration and degraded operation module is configured to calculate the optimal alternative path through a multi-objective optimization algorithm, execute the reconfiguration according to a safe sequence, realize the degraded operation of the system to maintain core functions, and generate detailed work orders to notify personnel for maintenance.

[0064] In summary, due to the adoption of the above technical solution, the beneficial effects of the present invention are:

[0065] 1. This invention achieves accurate identification of fault signals through a fault coefficient determination mechanism, effectively solving the problem of false alarms in traditional systems. The system integrates abnormal values ​​of operating conditions and the asynchronous rate of strongly correlated nodes, quantifies the fault coefficient with the area of ​​a triangle, and verifies the authenticity of the signal from two dimensions: static operating condition adaptation and dynamic group collaboration. It can accurately distinguish between sensor faults and actual equipment faults, avoiding unnecessary downtime or reconfiguration.

[0066] 2. This invention supports multi-scenario detection through a topology self-discovery module, including initial installation, post-maintenance, periodic, and status-triggered detection. It can automatically adapt to topology changes such as equipment replacement and pipeline modification without manual intervention, greatly improving system scalability and maintenance efficiency. Based on personalized performance baselines for detailed operating conditions, combined with equipment rotation, load balancing, and energy-saving control strategies, the system can dynamically adjust operating parameters to balance equipment wear, avoid branch overload, and reduce pipeline resistance loss and energy consumption. Attached Figure Description

[0067] Further details, features, and advantages of this application are disclosed in the following description of exemplary embodiments in conjunction with the accompanying drawings, in which:

[0068] Figure 1 This is a system structure diagram of the present invention. Detailed Implementation

[0069] Several embodiments of this application will now be described in more detail with reference to the accompanying drawings to enable those skilled in the art to implement this application. This application may be embodied in many different forms and for various purposes and should not be limited to the embodiments set forth herein. These embodiments are provided to make this application thorough and complete, and to fully convey the scope of this application to those skilled in the art. The embodiments described do not limit this application.

[0070] Unless otherwise defined, all terms used herein (including technical and scientific terms) shall have the same meaning as commonly understood by one of ordinary skill in the art to which this application pertains. It will be further understood that terms such as those defined in commonly used dictionaries shall be interpreted as having a meaning consistent with their meaning in the relevant field and / or the context of this specification, and shall not be interpreted in an idealized or overly formal sense unless expressly defined herein.

[0071] Example 1

[0072] Its specific implementation method is combined with the appendix Figure 1 Please provide a detailed explanation.

[0073] Appendix Figure 1 The block diagram of the control system for rail transit electromechanical equipment based on the Internet of Things provided in the embodiments of the present invention shows the connection relationship between the topology self-discovery and map building module and the dynamic control reconstruction and degraded operation module, and marks the main functional interaction flow of each module.

[0074] In this embodiment, it includes:

[0075] The topology self-discovery and map building module is configured to collect sensor data through standardized detection sequences, analyze device correlations through multi-algorithm processing, and generate a highly reliable structured topology model.

[0076] Specifically, it includes:

[0077] Triggering conditions:

[0078] Initial installation: After the system is powered on for the first time, the operating system and drivers are initialized (takes about 3-5 minutes), and the topology discovery is automatically triggered without manual intervention;

[0079] Alternatively, maintenance personnel can trigger the topology reconfiguration by entering an access password and clicking the topology reconfiguration button via an industrial touchscreen or remote monitoring platform. Before triggering, the system automatically detects the current device status and allows startup only when all actuators are in standby / off state, thus avoiding faults caused by detection during operation.

[0080] By default, it is automatically triggered quarterly (the cycle can be adjusted through backend configuration, ranging from 1 to 12 months); it also supports status-triggered supplementation: when the system detects a change in the device connection status (such as a new device being connected, or an existing device being offline for more than 10 minutes), it automatically starts incremental topology detection (only detecting the changed parts, shortening the detection time).

[0081] Detection execution process:

[0082] Preparation before detection: The central controller first sends a command to put all actuators (water pumps, solenoid valves) into the initial state (water pumps stop, solenoid valves close), and calibrates all sensors (flow, pressure, level sensors) to eliminate zero drift (calibration takes about 30 seconds, and the calibration standard is ±0.5% of the sensor range).

[0083] The detection sequence design principle is to execute the detection signals of different actuators in the following order: valves first, then pumps; branch lines first, then main lines; and near ends first, then far ends, in order to avoid mutual interference.

[0084] Detailed detection example (using a rail transit drainage system as an example):

[0085] Initial state: All 12 solenoid valves (numbered V1-V12) are closed, all 4 water pumps (numbered P1-P4) are stopped, and all sensors (8 flow sensors F1-F8, 6 pressure sensors Pr1-Pr6, and 4 level sensors L1-L4) enter high-frequency acquisition mode (sampling frequency 10Hz).

[0086] Solenoid valve V1 detection: Send an opening command to V1 for 2 seconds, and record the value changes of all sensors within 0-5 seconds (the first 2 seconds are the V1 opening phase, and the last 3 seconds are the recovery phase).

[0087] Data judgment criteria: If the value of flow sensor F1 increases from 0 to ≥5L / s (threshold is configurable), and other flow sensors do not show significant changes (change amount ≤0.5L / s), it is determined that V1 and F1 are directly connected;

[0088] The pressure sensor Pr1 value fluctuated from 0.3MPa to 0.35MPa and then recovered, indicating that there is a hydraulic connection between the branch where V1 is located and the main pipe monitored by Pr1.

[0089] Solenoid valve V2-V12 detection: Repeat the operation of the solenoid valve V1 detection step, with a 5-second interval between each solenoid valve detection, and ensure that the system status is fully restored before proceeding to the next detection.

[0090] Pump P1 detection: Send a command to P1 to start the frequency converter to 30Hz and run for 3 seconds (low frequency operation to avoid impact), and record all sensor data within 0-8 seconds;

[0091] Inlet side judgment: The value of level sensor L1 drops from 1.2m to 1.15m (level drop rate ≥ 0.01m / s), and the value of pressure sensor Pr2 drops from 0.2MPa to 0.18MPa. It is determined that the pipes containing L1 and Pr2 are the inlet path of P1.

[0092] Outlet side judgment: The flow sensor F2 and F3 values ​​rise to ≥8L / s simultaneously, and the pressure sensor Pr3 value rises to 0.4MPa. It is determined that the pipes where F2, F3, and Pr3 are located are the outlet paths of P1, and P1 serves the two drainage areas covered by F2 and F3.

[0093] Pump P2-P4 detection: Repeat the operation of pump P1 detection, with a 10-second interval between each pump detection to avoid the superposition of water flow disturbances affecting data accuracy.

[0094] Anomaly Handling: If the sensor does not change as expected during a probe (e.g., no flow sensor response after V3 is enabled), the system will automatically repeat the probe twice; if there is still no response, the connection between V3 and the sensor will be marked as abnormal, and subsequent probes will continue. After the probe is completed, the abnormal node will be marked on the topology map.

[0095] The collected sensor data is filtered using a moving average filtering method (window size of 5 sampling points) to remove spike noise caused by electromagnetic interference; at the same time, outliers exceeding ±10% of the sensor's range are removed to ensure data validity.

[0096] Details of the correlation algorithm:

[0097] Time-series correlation analysis: Calculate the Pearson correlation coefficient between actuator actions (such as V1 being turned on) and sensor value changes, and determine the correlation strength based on the Pearson correlation coefficient;

[0098] Pearson correlation coefficient: |r|≥0.8 indicates a strong association, 0.5≤|r|<0.8 indicates a weak association, and |r|<0.5 indicates no association.

[0099] The core of calculating the Pearson correlation coefficient is to quantify the degree of linear co-change between two variables. The value range is [-1, 1]. The larger the absolute value, the stronger the linear correlation. The calculation needs to be based on paired data of actuator action timing and sensor value change timing.

[0100] Detailed calculation steps (taking V1 activation and F1 flow change as an example):

[0101] Determine the time series data sequence of two variables:

[0102] Variable X (Actuator Action Timing): Based on the activation of V1, extract timing data before and after the action (e.g., 0-5 seconds, sampling frequency 10Hz, total 50 data points). Use 1 to indicate that the actuator action is effective (e.g., within 2 seconds of V1 activation), and 0 to indicate no action (2 seconds before activation, 3 seconds after activation), forming the X sequence. ;

[0103] Variable Y (sensor value change): Simultaneously capture the flow rate data of F1 within 0~5 seconds to directly form the Y sequence. Each data point corresponds one-to-one with the timestamp of the X sequence;

[0104] Calculate the mean of sequence X. ;

[0105] Calculate the mean of the Y sequence ;

[0106] Calculate covariance ;

[0107] Covariance calculation logic: Calculate the deviation of each data point from the mean of the corresponding sequence for each pair, multiply them, sum them, and then divide by (n-1) (the sample covariance needs to be adjusted for degrees of freedom).

[0108] Calculate the standard deviation of the two sequences to obtain , ;

[0109] Finally, the Pearson correlation coefficient was calculated. ;

[0110] The calculation process of the Pearson correlation coefficient is a direct reference to existing technology and will not be elaborated here.

[0111] Amplitude change threshold method: Set a minimum threshold for sensor value changes (e.g., flow rate ≥ 1 L / s, pressure ≥ 0.02 MPa, liquid level ≥ 0.01 m). Only when the change exceeds the threshold and matches the timing of the actuator action is it determined to be associated.

[0112] The topology graph data model structure includes:

[0113] Node table: Includes node ID (e.g., V1, P1, F1), equipment type (solenoid valve / pump / flow sensor), installation location (e.g., section 3 of track drainage branch 1), equipment parameters (e.g., pump rated power 5.5kW, solenoid valve diameter DN50), and health status (initially normal).

[0114] Edge table: contains edge ID, starting node ID, ending node ID, connection type (physical connection / hydraulic association), association strength (strong / weak), fluid flow direction (e.g., V1 to F1), and path length (calculated based on installation location, in meters).

[0115] Topology map storage and visualization: The structured data model is stored in a relational database (such as MySQL), supporting SQL query for relationships; at the same time, a visual topology map (SVG format) is generated and displayed on the monitoring platform. Nodes are distinguished by different colors to identify device types (red = water pump, blue = solenoid valve, green = sensor), edges are represented by solid lines to indicate strong relationships and dashed lines to indicate weak relationships, and flow directions are marked with arrows.

[0116] The routine monitoring and performance learning module is configured to optimize control strategies based on topology graphs, continuously collect equipment operation data, and establish and dynamically update personalized performance baselines that fit the working conditions.

[0117] Specifically, it includes:

[0118] Topology-based optimization control:

[0119] Control strategies include:

[0120] Equipment rotation strategy: Based on the same functional equipment group in the topology diagram (such as P1 and P2 both serving drainage area A), the equipment is rotated and started according to the running time. It is switched every 8 hours. When switching, the standby pump is started first, and the current pump is stopped after it has been running stably for a preset time (30 seconds) to avoid drainage interruption. At the same time, the cumulative running time of each equipment is recorded, and the deviation does not exceed 1 hour.

[0121] Load balancing strategy: Analyze the flow distribution of each branch through the topology diagram. When the flow of a branch is greater than or equal to the preset percentage (80%) of the rated flow (e.g., F1 detects a flow of 16L / s, rated 20L / s), the backup solenoid valve in the same area (e.g., V4) is automatically opened to divert the flow to the adjacent branch (e.g., the branch where F5 is located) to avoid overload of a single branch.

[0122] Energy-saving control strategy: Calculate the optimal operating combination by combining the path length and equipment parameters in the topology diagram; for example, when draining water to distant areas, prioritize starting high-power water pumps (such as P3, 7.5kW) and short-path valves to reduce pipeline resistance loss; when draining water to nearby areas, start low-power water pumps (such as P1, 5.5kW) to reduce energy consumption.

[0123] Based on real-time flow and pressure data, dynamically adjust the pump frequency converter (adjustment step size 0.5Hz, response time ≤2 seconds) and the solenoid valve switching status to ensure that the system operates in the optimal range where the flow meets the demand and energy consumption is minimized.

[0124] The control log records every control action (such as pump start-up, valve switching, frequency adjustment), including action time, triggering conditions (such as F1 flow rate of 18L / s, V4 diversion start), and execution result (such as V4 start-up successful, F1 flow rate drops to 12L / s). The log is kept for 1 year.

[0125] This also includes performance baseline learning:

[0126] The data collection dimensions and frequency are as follows:

[0127] Water pump: starting current (sampling frequency 5Hz, recording data for 10 seconds from start-up to steady state), steady-state operating current (1Hz), operating frequency (1Hz), vibration acceleration (3-axis, 10Hz), inlet and outlet pressure (1Hz), operating noise (1Hz, measurement range 40-100dB).

[0128] Solenoid valve: opening time (the time from the issuance of the command to the feedback of full opening, accurate to 10ms), closing time (same as opening time), coil current (1Hz), pressure difference across the valve (1Hz).

[0129] Sensor: Measured value (1Hz), measurement error (calibrated daily, error value recorded), response time (tested monthly, time from change of physical quantity to data output recorded);

[0130] The baseline establishment method is as follows:

[0131] Collect a preset number (at least 100 sets) of valid data (such as the current curve of 100 normal starts of the water pump) under normal operating conditions, and the data covers different operating loads (such as steady-state data of the water pump at 20Hz, 30Hz and 50Hz frequencies).

[0132] A baseline range is generated using statistical methods. For continuous data (such as operating current), the preset value of the sample (mean ± 2 times the standard deviation) is taken as the normal range. For time-series data (such as starting current curve), the average curve of the sample is taken as the reference curve, with an allowable amplitude deviation of ±15%.

[0133] Baselines are established according to equipment type and operating conditions (such as light load and heavy load for water pumps). For example, the steady-state current baseline of P1 water pump is 8-10A at 30Hz and 15-18A at 50Hz.

[0134] Baseline update mechanism:

[0135] Regular updates: The baseline is automatically updated every 3 months, incorporating the latest normal operation data and removing outdated data (more than 1 year old) to ensure that the baseline matches the current state of the equipment.

[0136] Triggered Update: After the equipment has undergone major repairs or replacement of key components (such as replacing the bearings of the water pump), the system will automatically trigger a baseline reset and re-collect 100 sets of data to establish a new baseline.

[0137] Baseline anomaly warning: If more than 30% of the newly collected normal data exceeds the current baseline range, the system will issue a baseline update warning, reminding maintenance personnel to manually trigger the update after confirming the equipment status.

[0138] The fault diagnosis and reconfiguration trigger module is configured to trigger fault signal diagnosis to obtain fault coefficients when the controller receives a fault feedback signal from the equipment, or when each real-time data is compared with the performance baseline under the corresponding operating condition and the comparison result meets the abnormal conditions. Based on the fault coefficients, it is determined whether the corresponding fault signal is erroneous. If the fault signal is erroneous, the fault type is determined and the scope of influence is analyzed, and the reconfiguration process is triggered.

[0139] Specifically, it includes:

[0140] Hardware fault detection:

[0141] The controller directly receives fault feedback signals from the equipment, including overcurrent protection signals (pump motor current exceeds 150% of the rated value), short circuit alarms (coil short circuit, feedback resistance ≤1Ω), mechanical jamming signals (no full open / full close feedback received 3 seconds after the solenoid valve command is issued), and power supply failures (equipment power supply voltage is lower than 85% of the rated value).

[0142] After a hard fault signal is triggered, the controller records the fault type and occurrence time within 50ms and immediately cuts off the control signal of the faulty device to prevent the fault from escalating.

[0143] Soft fault / performance degradation detection:

[0144] Single-parameter judgment: Compare real-time data with the performance baseline under the corresponding operating conditions. Data is considered abnormal if it meets the following conditions:

[0145] The water pump operating current exceeds the baseline range for a preset maximum duration (5 seconds, e.g., baseline 8-10A, actual 11A for 5 seconds).

[0146] The solenoid valve opening time is extended by a preset percentage (more than 30%, e.g., 0.5 seconds for the baseline and 0.7 seconds for the actual time).

[0147] The fluctuation range of the sensor measurement value exceeds the preset percentage of the baseline (20%, such as a flow baseline of 10±1L / s, with an actual fluctuation range of 8-12L / s).

[0148] Multi-parameter fusion judgment: When a single parameter is abnormal but does not reach the threshold, but multiple related parameters show slight abnormalities at the same time (such as a 10% increase in pump vibration acceleration, a 5dB increase in operating noise, and an 8% increase in current), an abnormality index is calculated by a weighted fusion algorithm (weights are assigned according to the importance of the parameters, such as vibration weight 0.4, current weight 0.3, and noise weight 0.3). If the abnormality index is greater than the preset threshold, it is judged as performance degradation.

[0149] The process of obtaining the fault coefficient includes:

[0150] Based on sensors, the operating frequency of the water pump, the liquid level on the suction side, and the regional flow demand corresponding to the fault signal are obtained, the sensor baseline range for the corresponding working condition is determined, and the maximum and minimum values ​​under the corresponding working condition are determined.

[0151] Pump operating frequency: read in real time from the frequency converter;

[0152] Suction-side liquid level: from the topologically correlated liquid level sensor;

[0153] Area flow requirements: extracted from system control commands (e.g., area A needs to drain 15L / s), or calculated through topologically associated level sensors;

[0154] First, collect the real-time values ​​of these three parameters, and then match them to the unique corresponding sensor baseline range through the system's preset working condition-baseline mapping table.

[0155] Example: The data collected shows an operating frequency of 30Hz, a suction side liquid level of 1.2m, and a regional flow requirement of 15L / s. The baseline of the flow sensor F2 is matched to 8-12L / s, and the baseline of the current sensor is 8-10A.

[0156] Calculate the operating condition deviation rate:

[0157] Operating condition deviation rate = (Target sensor measured value - Baseline median under corresponding operating condition) ÷ Baseline median under corresponding operating condition × 100%;

[0158] Wherein, the baseline median = (the maximum baseline value under the corresponding working condition + the minimum baseline value under the corresponding working condition) ÷ 2;

[0159] A preset operating condition deviation rate threshold is set. If the operating condition deviation rate is greater than the operating condition deviation rate threshold, the absolute value of the difference between the two is calculated and recorded as an abnormal value of the operating condition deviation.

[0160] The parameters with the largest changes among the pump operating frequency, suction side liquid level, and regional flow demand are selected as reference conditions. The data changes of the corresponding sensors within a preset time period and the corresponding changes of the operating parameters are obtained for the reference conditions.

[0161] Sensor data change: Stable value after operating condition change - Stable value before operating condition change;

[0162] Changes in operating parameters: Only the change in the parameter that triggered the change is taken;

[0163] The operating condition response value is obtained by dividing the change in sensor data within a preset time period by the change in operating parameters.

[0164] Obtain the corresponding baseline operating condition response value range, match the operating condition response value with the operating condition response value range, mark the operating condition response value that is not within the operating condition response value range, calculate the difference between it and the operating condition response value range, take the absolute value, and record it as an abnormal operating condition response value.

[0165] After normalizing the abnormal values ​​of operating condition deviation and operating condition response, their values ​​are set to the range of [0,1]. Weighted summation is then performed to obtain the abnormal values ​​of operating condition matching.

[0166] Weighted summation process: Preset weighting factors for abnormal operating condition deviations and abnormal operating condition responses. Multiply the abnormal operating condition deviations and abnormal operating condition responses by their corresponding weighting factors and sum them to obtain the abnormal operating condition matching values.

[0167] A precise and comprehensive operating condition matching verification system was constructed. By collecting three key operating condition parameters—pump operating frequency, suction side liquid level, and regional flow demand—the baseline range of the sensor corresponding to the operating condition was accurately located. Then, the deviation rate of the operating condition was used to quantify the degree of deviation between the sensor's measured value and the baseline. Combined with the operating condition response value, the synergy of the data with changes in operating conditions was verified. This achieved dual-dimensional signal verification of static and dynamic aspects, reducing the probability of signal misjudgment caused by fluctuations in operating conditions and making the verification of the authenticity of abnormal signals more rigorous and convincing.

[0168] By using normalization and weighted summation quantization logic, outliers of operating condition deviation and outliers of operating condition response are integrated into unified operating condition matching outliers. This not only makes outlier data from different dimensions comparable, but also adapts the verification focus under different scenarios through weight configuration, providing a standardized and high-precision input basis for subsequent calculation of fault coefficients.

[0169] Its clear parameter acquisition methods, calculation steps, and example descriptions transform abstract signal verification logic into concrete operations that can be implemented in engineering, which not only improves the reliability of system fault diagnosis but also reduces the debugging difficulty in actual deployment, perfectly adapting to the high reliability requirements of rail transit electromechanical equipment under complex operating conditions.

[0170] Starting with the faulty water pump corresponding to the sensor fault signal, the topology graph is traversed using the DFS algorithm to filter out all sensor nodes that are strongly correlated with the water pump, and sensors with a Pearson correlation coefficient greater than or equal to 0.8 are selected.

[0171] Acquire the node signal values ​​monitored by each sensor node strongly associated with the water pump, and preset the allowable fluctuation range of the node signal values;

[0172] If the allowable fluctuation range of the preset node signal value covers the sensor fault signal, then the sensor corresponding to the allowable fluctuation range of the preset node signal value is marked as a synchronization abnormal sensor.

[0173] After obtaining all the synchronous abnormal sensors, divide by the number of sensor nodes strongly associated with the water pump to get the synchronization rate, and subtract the synchronization rate from 1 to get the asynchronous rate.

[0174] Based on the strong correlation established in the topology self-discovery phase, the DFS algorithm is used to accurately screen out sensor nodes that are highly correlated with the faulty water pump (Pearson correlation coefficient ≥ 0.8), and the proportion of synchronously abnormal sensors is used to determine whether the signal abnormality is a group coordination phenomenon.

[0175] If only a single sensor malfunctions, the asynchrony rate will increase significantly, directly indicating a fault in the sensor itself. If most strongly correlated sensors malfunction synchronously, it confirms that the signal is a true reflection of a water pump malfunction. This group verification based on device correlation logic elevates the judgment of signal authenticity from the individual level to the system level, greatly improving the anti-interference capability and accuracy of fault diagnosis, and avoiding false triggering of reconstruction due to drift or damage of a single sensor.

[0176] By quantifying the synchronization rate and asynchrony rate, the abstract abnormal coordination of related nodes is transformed into intuitive and comparable numerical indicators, providing key input for the subsequent construction of fault coefficients.

[0177] Its clear criteria for selecting strongly correlated nodes and rules for judging synchronization anomalies ensure the consistency and repeatability of the verification logic, which not only connects with the previous results of the topology self-discovery module, but also provides a standardized quantitative basis for fault diagnosis.

[0178] After normalizing the abnormal values ​​of the operating condition matching and the asynchronous rate, their value range is [0,1]. The abnormal values ​​of the operating condition matching and the asynchronous rate are respectively used as the base and the height perpendicular to the base of the triangle to construct the triangle, and the area of ​​the triangle is calculated and denoted as the fault coefficient.

[0179] A preset fault coefficient threshold is set. If the fault coefficient is greater than the fault coefficient threshold, the sensor corresponding to the fault coefficient is determined to be faulty, and the fault signal is incorrect.

[0180] By eliminating the dimensional differences between outliers in operating condition matching and asynchronous rate through normalization, and then integrating the two key indicators into a single fault coefficient using the triangle area formula, the dual verification logic of operating condition adaptability and group synergy is preserved, while transforming the abstract multi-dimensional verification into intuitive and comparable quantitative values.

[0181] This design upgrades the judgment of fault signals from multi-dimensional qualitative description to single-index quantitative judgment. By setting a threshold, it can clearly distinguish between sensor faults and actual pump faults, improving the accuracy and consistency of fault diagnosis and avoiding decision hesitation or misjudgment caused by conflicting results from multiple dimensions.

[0182] It also includes an analysis of the scope of impact:

[0183] The Depth-First Search (DFS) algorithm is used to start with the faulty device, traverse all nodes in the topology graph that are strongly associated with the device to determine the direct impact range; then traverse the weakly associated nodes to determine the indirect impact range.

[0184] The analytical dimensions and standards include:

[0185] At the equipment level: Is there a backup device to replace the function of the faulty device (e.g., if P1 fails, check the topology diagram to see if there are other water pumps serving the same area as P1);

[0186] At the regional level: Is there any other drainage path covering the drainage area (e.g., area A) corresponding to the faulty equipment (check the topology map to see if area A is connected to other branches);

[0187] Functional level: Does the fault cause the interruption of core functions (such as whether it affects key functions such as track flood control and drainage, equipment cooling and drainage);

[0188] Based on the impact level, it includes minor impact, moderate impact, and severe impact;

[0189] Minor impact: Only a single non-core branch is affected; core functions are uninterrupted; backup paths can fully cover the impact.

[0190] General impact: Multiple branches or a single core branch are affected, core functions are partially restricted, and backup paths can provide partial coverage.

[0191] Severe impact: Multiple core branches are affected, core functions are at risk of interruption, and emergency paths need to be activated immediately;

[0192] The analysis results are output as an impact scope analysis report, which includes the faulty device ID, fault type, list of directly affected nodes (such as F2 and F3 sensors, drainage area A), list of indirectly affected nodes (such as Pr3 pressure sensor), impact level, and degree of impact on core functions, providing clear targets for subsequent reconstruction.

[0193] The dynamic control reconfiguration and degradation operation module is configured to calculate the optimal alternative path using a multi-objective optimization algorithm, execute reconfiguration according to a safe sequence, achieve system degradation operation to maintain core functions, and generate detailed work orders to notify personnel for maintenance. Specifically, this includes:

[0194] The alternative path calculation process is as follows:

[0195] Calculate the input parameters:

[0196] Topology graph data: the latest node status (healthy / faulty), edge connectivity and association strength.

[0197] Equipment parameters: Rated power, flow rate, and pressure range of the standby equipment, and its current health status (e.g., P2 health rate 90%, P3 health rate 85%).

[0198] Operational data: Current flow demand in each area (e.g., area A currently needs to drain 15L / s), and pipeline pressure distribution (e.g., main pipeline pressure 0.3MPa).

[0199] Constraints: Standby equipment must not be overloaded (operating current ≤ 110% of rated current), pipeline pressure must not exceed the safe range (0.2-0.5MPa), and drainage needs of the core area must be met first.

[0200] Example of calculation process (P1 fault, area A needs to drain 15L / s):

[0201] Standby water pumps with a health status ≥80% and a rated flow rate ≥15L / s were selected from the topology diagram: P2 (rated flow rate 20L / s, health status 90%) and P3 (rated flow rate 25L / s, health status 85%).

[0202] Filtering alternative branches connected to region A: By traversing the topology graph, it is found that P2 can be connected via paths from V5, V6, F2, and region A, and P3 can be connected via paths from V7, F3, and region A.

[0203] Calculate the objective function value for each path:

[0204] P2 path: Functional coverage completeness 100% (20L / s≥15L / s), load rate 75% (15 / 20), energy consumption 8kW / h, objective function value = 1×0.5+(1-0.75)×0.3+(1-8 / 10)×0.2=0.5+0.075+0.04=0.615.

[0205] The values ​​0.5, 0.3, and 0.2 are the weighting factors for each parameter, which are preset values ​​and are used for calculation reference only.

[0206] P3 path: 100% functional coverage, 60% load rate (15 / 25), energy consumption 10kW / h, objective function value = 1×0.5+(1-0.6)×0.3+(1-10 / 12)×0.2=0.5+0.12+0.033=0.653.

[0207] Choose the P2 path with a higher objective function value. The optimal alternative is to enable V5, V6 and start P2 (frequency conversion to 35Hz, output flow rate 15L / s).

[0208] Calculation anomaly handling: If no single backup path meets the functional coverage requirements, automatically calculate a multi-device collaborative alternative solution (such as P2 and P4 operating together); if all solutions fail to meet the core functional requirements, directly trigger the highest level emergency alarm and notify personnel for emergency handling.

[0209] The safe execution refactoring sequence is as follows:

[0210] Sequence planning principle: Execute in the order of first establishing the path, then starting the equipment, and finally adjusting the parameters. Each step has a feedback confirmation step to ensure the execution effect.

[0211] Detailed execution steps (taking P2 path replacing P1 as an example):

[0212] Path preprocessing: Send pre-opening commands to V5 and V6 to raise the valve drive voltage to 50% of the rated value, in preparation for opening (to avoid instantaneous high pressure impact on the valve).

[0213] Open valve: Send a full open command, and V5 and V6 will open simultaneously. The opening speed is controlled at 0.2m / s (for gate valves) to avoid water flow impact.

[0214] Valve status confirmation: Wait 2 seconds (the longest time required for the valve to be fully open and the redundancy time), receive the full-open feedback signals from V5 and V6, and simultaneously monitor the values ​​of Pr4 (pressure sensor after V5) and Pr5 (pressure sensor after V6). If the pressure is stable (fluctuation ≤ 0.02MPa), the valve is considered to have been successfully opened.

[0215] Start the standby pump: Send a frequency converter start command to P2, with an initial frequency of 15Hz (low frequency start to avoid current surge), and a start time of 3 seconds (from 15Hz to the target frequency of 35Hz).

[0216] Pump operation confirmation: Monitor the operating current of P2 (≤12A, 120% of the rated current of 10A), vibration acceleration, and inlet and outlet pressures (inlet ≥0.18MPa, outlet ≤0.45MPa). If there are no abnormalities for 3 seconds, the pump is considered to have started successfully.

[0217] Flow rate adjustment: Based on the real-time data of F2 (A-zone flow sensor), dynamically adjust the operating frequency of P2 by 0.5Hz every 0.5 seconds until the F2 value stabilizes at 15L / s±0.5L / s, thus completing parameter calibration.

[0218] System status confirmation: Monitor the value of the liquid level sensor L1 in area A and ensure that it is stable within the safe range (0.8-1.2m). At the same time, check that the sensor data in other areas (such as F4 and F5) show no abnormal fluctuations. If the data shows that the reconstruction sequence was executed successfully, the system status is confirmed.

[0219] Fault handling: If no confirmation feedback is received for a certain step (e.g., no full-on signal is received 3 seconds after V5 is enabled), the system will automatically retry 2 times; if it still fails, it will switch to the backup solution (e.g., enable the P3 path); if all backup solutions fail, the reconstruction will be stopped immediately, the current state will be maintained and a reconstruction failure alarm will be sent to notify personnel to intervene.

[0220] The process of system downgrading and personnel intervention is as follows:

[0221] Degradation operation parameter limitations:

[0222] Equipment operation restrictions: The operating frequency of the standby equipment is ≤110% of the rated frequency (e.g., P2's rated frequency is 50Hz, and it is ≤55Hz when operating under degraded conditions), the operating current is ≤120% of the rated current, and the cumulative operating time is ≤72 hours (after which a forced shutdown will be automatically triggered to avoid equipment damage).

[0223] Functional priority: Core functions (such as track flood control and drainage) should be prioritized, while non-core functions (such as greening irrigation and drainage) can be suspended or have their flow reduced (≤50% of rated flow).

[0224] Energy consumption control: During degraded operation, energy consumption is allowed to be 30% higher than normal operation, but energy consumption data needs to be monitored in real time. If it exceeds 50%, the operating parameters should be adjusted (such as appropriately reducing traffic in non-core areas).

[0225] Repair work order details:

[0226] Basic fault information: Faulty device ID (P1), device model (ISG50-160), installation location (drainage pump pit in area A of track 1), fault occurrence time (accurate to the second), fault type (operating current is continuously high, determined to be impeller blockage).

[0227] Emergency response information: Alternative solutions activated (V5 and V6 enabled, P2 running, frequency 35Hz), current system operating status (Area A drainage is normal, flow rate 15L / s, energy consumption 8kW / h), degraded operation restrictions (P2 cumulative running time ≤72 hours).

[0228] Maintenance recommendations: Recommended maintenance time (within 24 hours), required tools (impeller disassembly tools, pipe cleaning equipment), maintenance steps (shutdown, disassembly of pump body, cleaning of impeller blockage, reassembly, testing), safety precautions (wait 5 minutes after power off to ensure the pump body has completely stopped).

[0229] Personnel notification method:

[0230] Tiered notification: Minor impact faults are only notified to on-site maintenance personnel (via industrial SMS and platform messages); moderate impact faults are notified to both on-site maintenance personnel and team leaders; severe impact faults are notified to on-site personnel, team leaders, and operations managers, and also trigger on-site audible and visual alarms.

[0231] Notification frequency: Before maintenance is confirmed, a notification will be sent every 30 minutes; after maintenance personnel confirm receipt, the system operating status will be updated every 2 hours.

[0232] Post-repair recovery process:

[0233] After maintenance personnel complete the repair, they click "Repair Complete" on the monitoring platform, which triggers topology auto-discovery (incremental detection, only detecting P1 and related nodes).

[0234] The system regenerates the topology map, confirms that the state of P1 has returned to normal, and that the association with the relevant nodes is correct.

[0235] Execute the recovery sequence: First, reduce the P2 frequency to 15Hz and then stop the machine, shut down V5 and V6, then start P1 and adjust it to the normal operating parameters.

[0236] After monitoring P1 and related sensor data for 30 seconds and confirming that there are no abnormalities, the system exits the degraded operation mode and resumes normal control logic.

[0237] The above formulas are all dimensionless calculations. The formulas are derived from software simulations based on a large amount of collected data to obtain the most recent real-world results. The preset parameters in the formulas are set by those skilled in the art according to the actual situation.

[0238] The foregoing has only described certain exemplary embodiments of the present invention by way of illustration. Undoubtedly, those skilled in the art can modify the described embodiments in various ways without departing from the spirit and scope of the present invention. Therefore, the foregoing drawings and descriptions are illustrative in nature and should not be construed as limiting the scope of protection of the claims of the present invention.

[0239] It should be noted that, in this document, the use of relational terms such as "first" and "second" is merely for distinguishing one entity or operation from another, and does 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. Without further limitations, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes the element.

[0240] It should be understood that in the various embodiments of this application, the order of the above-mentioned processes does not imply the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of this application.

[0241] Those skilled in the art will recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0242] Those skilled in the art will understand that, for the sake of convenience and brevity, the specific working processes of the systems, devices, and units described above can be referred to the corresponding processes in the foregoing method embodiments, and will not be repeated here.

[0243] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0244] In addition, the functional units in the various embodiments of this application can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit.

[0245] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

[0246] The foregoing has only described certain exemplary embodiments of the present invention by way of illustration. Undoubtedly, those skilled in the art can modify the described embodiments in various ways without departing from the spirit and scope of the present invention. Therefore, the foregoing drawings and descriptions are illustrative in nature and should not be construed as limiting the scope of protection of the claims of the present invention.

Claims

1. A rail transit electromechanical equipment control system based on the Internet of Things, characterized in that, include: The topology self-discovery and map building module is configured to collect sensor data through standardized detection sequences, analyze device correlations through multi-algorithm processing, and generate a highly reliable structured topology model. The routine monitoring and performance learning module is configured to optimize control strategies based on topology graphs, continuously collect equipment operation data, and establish performance baselines. The fault diagnosis and reconfiguration trigger module is configured to trigger fault signal diagnosis to obtain fault coefficients when the controller receives a fault feedback signal from the equipment, or when the comparison results of each real-time data with the performance baseline under the corresponding operating conditions meet the abnormal conditions. Based on the fault coefficients, it is determined whether the corresponding fault signal is erroneous. If the fault signal is erroneous, the fault type is determined and the scope of influence is analyzed, and the reconfiguration process is triggered. The process of obtaining the fault coefficient includes: After analyzing the sensors, abnormal values ​​of operating condition deviation and abnormal values ​​of operating condition response are obtained. These are then weighted and summed to obtain abnormal values ​​of operating condition matching. After analyzing the number of synchronous abnormal sensors and sensor nodes strongly associated with the water pump, the asynchronous rate was obtained; The fault coefficient is obtained by combining abnormal values ​​of operating conditions with the asynchronous rate.

2. The IoT-based rail transit electromechanical equipment control system according to claim 1, characterized in that, The topology self-discovery and map building module specifically includes: After the system is powered on for the first time, it completes the initialization of the operating system and drivers, and automatically triggers topology discovery. The central controller first sends instructions to put all actuators into their initial state and calibrate all sensors to eliminate zero drift; The collected sensor data is filtered to remove spike noise caused by electromagnetic interference; at the same time, outliers that exceed the sensor's measurement range are also removed. Calculate the Pearson correlation coefficient between actuator action and sensor value changes, and determine the correlation strength based on the Pearson correlation coefficient; Set a minimum threshold for sensor value changes; only when the change exceeds the threshold and matches the timing of actuator actions is it determined to be associated. The topology graph data model structure includes: Node table: contains node ID, device type, installation location, device parameters, and health status; Edge table: contains edge ID, starting node ID, ending node ID, connection type, association strength, fluid flow direction, and path length.

3. The IoT-based rail transit electromechanical equipment control system according to claim 1, characterized in that, The routine monitoring and performance learning module specifically includes: Based on the topology diagram, devices with the same function are started in rotation according to their runtime; at the same time, the cumulative runtime of each device is recorded. By analyzing the flow distribution of each branch through the topology diagram, when the flow of a branch is greater than or equal to a preset percentage of the rated flow, the backup solenoid valve in the same area is opened to divert the flow to the adjacent branch. Based on real-time flow and pressure data, the frequency of the water pump and the on / off status of the solenoid valve are dynamically adjusted to ensure that the system operates in the optimal range where the flow meets the demand and energy consumption is minimized.

4. The Internet of Things-based rail transit electromechanical equipment control system according to claim 3, characterized in that, This also includes performance baseline learning: The data collection dimensions and frequency are as follows: Water pump: starting current, steady-state operating current, operating frequency, vibration acceleration, inlet and outlet pressure, operating noise; Solenoid valve: opening time, closing time, coil current, pressure difference across the valve; Sensor: Measured value, measurement error, response time; The baseline establishment method is as follows: Collect a preset number of valid data points under normal operating conditions, and ensure that the data covers different operating loads; Statistical methods are used to generate baseline ranges. For continuous data, the preset values ​​of the samples are taken as the normal range; for time series data, the average curve of the samples is taken as the baseline curve. Baselines are established according to equipment type and operating conditions.

5. The Internet of Things-based rail transit electromechanical equipment control system according to claim 1, characterized in that, The fault diagnosis and refactoring trigger module specifically includes: The controller directly receives fault feedback signals from the equipment, including overcurrent protection signals, short circuit alarms, mechanical jamming signals, and power failure signals. Real-time data is compared with the performance baseline under the corresponding operating conditions, and is judged to be abnormal if it meets the following conditions: The water pump operating current exceeds the baseline range for a preset maximum duration. The solenoid valve opening time is extended by a preset percentage compared to the baseline. The sensor measurement value fluctuates by a preset percentage beyond the baseline.

6. The IoT-based rail transit electromechanical equipment control system according to claim 5, characterized in that, The process of obtaining the fault coefficient includes: Based on sensors, the operating frequency of the water pump, the suction side liquid level, and the regional flow demand corresponding to the fault signal are obtained, and the sensor baseline range for the corresponding operating condition is determined. Operating condition deviation rate = (Target sensor measured value - Baseline median under corresponding operating condition) ÷ Baseline median under corresponding operating condition × 100%; Baseline median = (maximum baseline value under the corresponding working condition + minimum baseline value under the corresponding working condition) ÷ 2; A preset operating condition deviation rate threshold is set. If the operating condition deviation rate is greater than the operating condition deviation rate threshold, the absolute value of the difference between the two is calculated and recorded as an abnormal value of the operating condition deviation. The parameters with the largest changes among the pump operating frequency, suction side liquid level, and regional flow demand are selected as reference conditions. The data changes of the corresponding sensors within a preset time period and the corresponding changes of the operating parameters are obtained for the reference conditions. The operating condition response value is obtained by dividing the change in sensor data within a preset time period by the change in operating parameters. Obtain the corresponding baseline operating condition response value range, match the operating condition response value with the operating condition response value range, mark the operating condition response value that is not within the operating condition response value range, calculate the difference between it and the operating condition response value range, take the absolute value, and record it as an abnormal operating condition response value. After normalizing the abnormal values ​​of operating condition deviation and operating condition response, their values ​​are set to the range of [0,1]. Weighted summation is then performed to obtain the abnormal values ​​of operating condition matching.

7. The IoT-based rail transit electromechanical equipment control system according to claim 6, characterized in that, Also includes: Starting with the faulty water pump corresponding to the sensor fault signal, traverse the topology graph and filter out all sensor nodes that are strongly associated with the water pump. Acquire the node signal values ​​monitored by each sensor node strongly associated with the water pump, and preset the allowable fluctuation range of the node signal values; If the allowable fluctuation range of the preset node signal value covers the sensor fault signal, then the sensor corresponding to the allowable fluctuation range of the preset node signal value is marked as a synchronization abnormal sensor. After obtaining all the synchronous abnormal sensors, divide by the number of sensor nodes strongly associated with the water pump to get the synchronization rate, and subtract the synchronization rate from 1 to get the asynchronous rate.

8. The Internet of Things-based rail transit electromechanical equipment control system according to claim 7, characterized in that, After normalizing the abnormal values ​​of the operating condition matching and the asynchronous rate, their value range is [0,1]. The abnormal values ​​of the operating condition matching and the asynchronous rate are respectively used as the base and the height perpendicular to the base of the triangle to construct the triangle, and the area of ​​the triangle is calculated and denoted as the fault coefficient. A preset fault coefficient threshold is set. If the fault coefficient is greater than the fault coefficient threshold, the sensor corresponding to the fault coefficient is determined to be faulty, and the fault signal is incorrect.

9. The Internet of Things-based rail transit electromechanical equipment control system according to claim 8, characterized in that, It also includes an analysis of the scope of impact: Starting with the faulty device, traverse all nodes in the topology graph that are strongly associated with the device to determine the direct impact range; then traverse the weakly associated nodes to determine the indirect impact range. The analysis dimensions include the device level, the region level, and the function level; Based on the impact level, it includes minor impact, moderate impact, and severe impact; The analysis results will be output as an impact range analysis report.

10. The Internet of Things-based rail transit electromechanical equipment control system according to claim 1, characterized in that, Also includes: The dynamic control reconfiguration and degraded operation module is configured to calculate the optimal alternative path through a multi-objective optimization algorithm, execute the reconfiguration according to a safe sequence, realize the degraded operation of the system to maintain core functions, and generate work orders to notify personnel for maintenance.

Images