Digital intelligent DC power supply operation and maintenance method, system and device, and medium

Abstract

The invention relates to the field of power electronics and power management, and discloses a digital-intelligent direct-current power supply operation and maintenance method, system and device and a medium, and the method comprises the following steps: obtaining direct-current power supply system operation parameters collected by a high-precision sensor; performing digital filtering processing on the operation parameters to generate pre-processing data; analyzing the preprocessed data based on a machine learning algorithm, and generating a storage battery charge state, a health state diagnosis result and system state data; dynamically adjusting output parameters of a charging module according to the system state data; based on the historical and real-time operation parameters, generating a fault early warning signal through a fault prediction model; and generating fault diagnosis and processing suggestions according to the fault type knowledge graph. According to the method, through Kalman filtering, the support vector machine, the ARIMA model and the RDF knowledge graph, operation parameters are monitored in real time, faults are predicted, accurate diagnosis suggestions are generated, and the system downtime risk is reduced.

Description

Technical Field

[0001] The present invention relates to the field of power electronics and power management, and in particular to a digital and intelligent direct current power supply operation and maintenance method, system, device and medium. Background Art

[0002] DC power supply systems are widely used in the fields of communications, electricity, rail transportation, etc., providing stable power supply for critical loads. Their operation and maintenance management is an important part of ensuring system reliability. In the existing technology, DC power supply operation and maintenance usually uses sensors to collect operating parameters (such as battery voltage, current, temperature and bus voltage), combined with threshold detection methods to monitor system status, identify AC input anomalies, DC bus voltage fluctuations or battery failures, etc. Some systems use simple statistical models or empirical rules for fault diagnosis, and transmit data to the monitoring terminal through a local display or communication interface (such as RS485). In addition, in terms of battery maintenance, existing technologies often evaluate capacity through regular charge and discharge tests, or use passive balancing circuits to adjust the voltage consistency of single cells to extend battery life.

[0003] However, existing technologies have significant deficiencies in fault prediction accuracy and system reliability. Traditional threshold detection methods rely on fixed parameters (such as voltage ±10%) and are difficult to adapt to dynamic changes in operating parameters, resulting in low fault prediction accuracy and high false alarm or missed alarm rates (prediction errors often exceed 10%). In addition, existing systems rely heavily on manual inspections and regular maintenance, resulting in delayed fault response, which can easily lead to unexpected downtime and reduced system availability. Especially under high load or extreme environments, downtime can extend to several hours, seriously affecting power supply reliability. Summary of the Invention

[0004] In order to remedy the above shortcomings, the present invention provides a digital DC power supply operation and maintenance method, system, device and medium, aiming to improve the problem that the traditional threshold detection method relies on fixed parameters and is difficult to adapt to the dynamic changes of operating parameters, resulting in low fault prediction accuracy.

[0005] In a first aspect, the present invention provides the following technical solution: a digitalized DC power supply operation and maintenance method, applied to a DC power supply system, with a monitoring device as the execution body, comprising the following steps: Acquire the DC power system operating parameters collected by high-precision sensors, including battery cell voltage, current, internal resistance, temperature, DC bus voltage, current, and feeder circuit voltage and current; Performing digital filtering on the operating parameters to generate pre-processed data; Analyze the preprocessed data based on a machine learning algorithm to generate battery state of charge, health status diagnosis results and system status data; Dynamically adjust output parameters of the charging module based on the system status data, wherein the output parameters are used to support demand-side response and peak-shaving; Generate fault warning signals through fault prediction models based on historical and real-time operating parameters; Generate fault diagnosis and handling suggestions based on the fault type knowledge graph; The system status data, fault warning signals and processing suggestions are displayed through a visual interface and uploaded to a remote terminal via Ethernet or CAN bus.

[0006] With this technical solution, the monitoring device acquires operating parameters through high-precision sensor interfaces deployed on the DC power system's battery pack, DC bus, and feeder circuit. The collected data is converted to digital signals via an analog-to-digital converter. Digital filtering employs a Kalman filter algorithm to eliminate noise interference in the voltage and current signals. Machine learning analysis, based on a support vector machine model, takes preprocessed voltage, current, internal resistance, and temperature data as input to generate state-of-charge (SOC) and state-of-health (SOH) diagnostic results. SOC is calculated using the voltage-current integration method, while SOH is assessed by combining internal resistance and capacity decay rate. System status data includes bus voltage stability and load power distribution. Based on this data, the monitoring device uses a PID algorithm to adjust the output voltage and current of the charging module, implementing demand-side response and prioritizing increased charging power during periods of low electricity demand. The fault prediction model utilizes time series analysis, combining historical data with real-time parameters to predict abnormal battery cell voltage or rectifier module overheating risks. A fault type knowledge graph, based on the RDF framework, stores the associations between fault characteristics and treatment recommendations, generating diagnostic recommendations through SPARQL queries. The visual interface uses an LCD display to display real-time parameters and alarm information. The data is uploaded to the remote operation and maintenance center via Ethernet (supporting IEC61850 protocol) or CAN bus.

[0007] Preferably, it also includes: The battery is desulfurized online through low-current pulse charge and discharge technology, and optimized battery performance data is generated by combining active and passive balancing. During peak and valley periods of electricity consumption, quantitative charging and discharging are performed based on the pre-processed data to generate calibrated battery capacity data for online capacity verification.

[0008] Preferably, it also includes: Based on the periodic change data of the feeding circuit current and power, the working status and health status evaluation results of the load equipment are generated; when the evaluation results indicate an abnormality, a load abnormality warning signal is generated and uploaded to the remote terminal via Ethernet or CAN bus.

[0009] Preferably, the fault warning signal includes at least one of the following: AC input overvoltage, undervoltage or phase loss; The DC bus or battery voltage is abnormal; The AC incoming line, busbar, battery or feeder switch trips; Rectifier module or lightning arrester failure; Abnormal battery charging or abnormal single cell voltage, internal resistance, and temperature.

[0010] In a second aspect, the present invention provides the following technical solutions: a digital and intelligent DC power supply operation and maintenance system, comprising: A data acquisition module is configured to acquire operating parameters of the DC power supply system through high-precision sensors. The operating parameters include battery cell voltage, current, internal resistance, temperature, DC bus voltage and current, and feeder circuit voltage and current. A data processing module is configured to perform digital filtering and machine learning analysis on the operating parameters to generate battery state of charge and health diagnosis results and system status data. a control module configured to dynamically adjust power supply output parameters according to the system status data and generate fault diagnosis and treatment suggestions; A communication module is configured to transmit the operating parameters, system status data and processing suggestions via Ethernet or CAN bus.

[0011] Preferably, the data processing module includes: a preprocessing unit configured to eliminate noise in the operating parameters by using a digital filtering technique to generate preprocessed data; The analysis unit is configured to analyze the preprocessed data based on a machine learning algorithm to generate SOC and SOH diagnostic results; the prediction unit is configured to generate a fault warning signal based on historical and real-time operating parameters.

[0012] Preferably, the control module includes: a power management unit configured to dynamically adjust the output voltage and current of the charging module according to the system status data to support peak shaving and valley filling; Desulfurization unit, configured to perform online desulfurization of the battery through low-current pulse charge and discharge technology, combining active and passive balancing to optimize battery performance; In a third aspect, the invention provides the following technical solution: a digital and intelligent DC power supply operation and maintenance device, comprising: processor; a memory storing instructions executable by the processor; When the processor executes the instructions, the steps of the above-mentioned digital DC power supply operation and maintenance method are implemented.

[0013] Preferably, the processor is further configured to: Receive remote control commands and perform remote charging and discharging, parameter configuration, or fault handling operations. The remote control commands comply with DL / T459, DL / T781, DL / T1074, and GB / T19826 power communication protocols; The operating parameters, system status data and fault warning signals are displayed through a visual interface.

[0014] In a fourth aspect, the present invention provides the following technical solution: a computer-readable storage medium storing a computer program, which, when executed by a processor, implements the steps of the above-mentioned digital DC power supply operation and maintenance method.

[0015] The present invention has the following beneficial effects: 1. In the present invention, Kalman filtering, support vector machine, ARIMA model and RDF knowledge graph are used to monitor operating parameters (voltage, current, internal resistance, temperature) in real time, predict faults (such as battery abnormalities and bus voltage fluctuations), generate accurate diagnostic suggestions, and reduce the risk of system downtime.

[0016] 2. In the present invention, through small current pulse charging and discharging and active / passive balancing, online desulfurization is achieved, internal resistance is reduced, and the consistency of single-cell voltage is optimized; quantitative charging and discharging during peak and valley periods calibrates capacity and extends battery life.

[0017] 3. In the present invention, based on the system status data (bus voltage stability and load power distribution), the charging module output is dynamically adjusted through PID control, the charging power is increased during the low electricity consumption period, and the power is reduced during the peak period, so as to achieve peak shaving and valley filling and optimize the load balance of the power grid.

[0018] 4. In the present invention, the periodic characteristics of the feed circuit current and power are extracted by Fourier transform, and the random forest algorithm is used to evaluate the load status (normal, overload, underload) and health status, generate early warning signals, and improve load management efficiency. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] Figure 1 This is a data processing flow chart of the digital and intelligent DC power supply operation and maintenance method, system, device and medium proposed in the present invention; Figure 2 This is a flow chart of the desulfurization and capacity control of the digital and intelligent DC power supply operation and maintenance method, system, device, and medium proposed in the present invention; Figure 3 This is a fault prediction and diagnosis flow chart for the digital and intelligent DC power supply operation and maintenance method, system, device, and medium proposed in the present invention; Figure 4 This is a system architecture diagram of the digital DC power supply operation and maintenance method, system, device and medium proposed in the present invention. DETAILED DESCRIPTION

[0020] The following will clearly and completely describe the technical solutions in the embodiments of the present invention with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, not all embodiments. All other embodiments obtained by ordinary technicians in this field based on the embodiments of the present invention without making any creative efforts shall fall within the scope of protection of the present invention.

[0021] Example 1 Reference Figure 1-Figure 3 In a first embodiment of the present invention, a digitalized DC power supply operation and maintenance method is provided, which is applied to a DC power supply system and is executed by a monitoring device, and includes the following steps: Acquire the DC power system operating parameters collected by high-precision sensors, including battery cell voltage, current, internal resistance, temperature, DC bus voltage, current, and feeder circuit voltage and current; Perform digital filtering on the operating parameters to generate pre-processed data; Analyze pre-processed data based on machine learning algorithms to generate battery state of charge, health status diagnosis results and system status data; Dynamically adjust the output parameters of the charging module based on system status data. The output parameters are used to support demand-side response and peak-shaving. Generate fault warning signals through fault prediction models based on historical and real-time operating parameters; Generate fault diagnosis and handling suggestions based on the fault type knowledge graph; System status data, fault warning signals and processing suggestions are displayed through a visual interface and uploaded to a remote terminal via Ethernet or CAN bus.

[0022] Specifically, the monitoring device obtains the operating parameters of the DC power supply system at a kHz sampling rate through a high-precision sensor interface. The sensors are deployed on the battery pack, DC bus and feeder circuit. The collected data include the battery cell voltage V i , current I i 、Internal resistance R i , temperature T i ,DC bus voltage V dc , current I dc , and the feed circuit voltage V f , current I f , converted into digital signals through analog-to-digital converters. Digital filtering uses the Kalman filter algorithm, and the state vector X k =[V k ,I k ] T Represents the estimated voltage and current at time k, and the observation vector Z k =[V m ,Im ] T Represents the sensor measurement value, and the state transition equation is X k =AX k-1 +w k , the observation equation is Z k =HX k +v k , where A and H are unit matrices, w k and v k The noise is Gaussian white noise, with covariance matrix Q = 0.01 (process noise), R = 0.05 (measurement noise), and the noise of the output pre-processed data is less than 0.01%. The machine learning analysis uses the support vector machine (SVM) model, with the input feature vector X = [V i ,I i ,R i ,T i ], label y is the state of charge (SOC, 0-100%) or state of health (SOH, 0-100%), a radial basis function (RBF) kernel is used, kernel parameter γ = 0.1, regularization parameter C = 1.0, the training data set contains 1000 sets of historical parameters, SOC error < 5%, SOH error < 3%, and system status data (bus voltage stability and load power distribution) are generated. Dynamic adjustment uses the PID control algorithm, input error e(t) = V ref -V out , where V ref is the target voltage, V out is the actual output voltage, which controls the output Proportional coefficient K p =0.5, integral coefficient K i =0.1, differential coefficient K d =0.05 adjusts the voltage range to ±10V and the current range to ±5A, supports demand-side response, and increases charging power during off-peak periods. Fault prediction uses the ARIMA model, with the input time series Y t =[V t ,R t ,T t ], parameters p = 2 (autoregressive order), d = 1 (differential order), q = 1 (moving average order), predict the probability of failure (such as internal resistance abnormality > 80%). The fault type knowledge graph is based on the Resource Description Framework (RDF), and the nodes represent the fault characteristics (such as internal resistance > 5mΩ ° ,Edges represent processing suggestions (e.g., “replace monomer”), and diagnostic ,suggestions are generated through SPARQL queries.,The visual interface uses an LCD display to display real-time ,parameters, status data, and alarm information.,The data is uploaded to the remote operation and ,maintenance center via Ethernet (supporting EC61850 protocol) or CAN bus, ,and the data packet format is JSON.

[0023] Also includes: The battery is desulfurized online through low-current pulse charge and discharge technology, and optimized battery performance data is generated by combining active and passive balancing. During peak and valley periods of electricity consumption, quantitative charging and discharging are performed based on pre-processed data to generate calibrated battery capacity data for online capacity verification.

[0024] Specifically, the monitoring device performs online desulfurization through small current pulse charge and discharge technology, generating frequency f = 10-100Hz, duty cycle D = 50, current amplitude I p =0.05C pulse signal (C is the rated capacity of the battery), triggering the decomposition of lead sulfate crystals, the decomposition efficiency Where ΔR is the internal resistance reduction, R0 is the initial internal resistance. Active balancing transfers power through the DC-DC converter, and the transfer power P t =V i I t , where V i is the cell voltage, I t is the transfer current (0.01C); passive balancing is achieved through the parallel resistor (R e =10Ω) dissipates excess power, optimizes the cell voltage consistency, and generates performance data reflecting the battery health status. During the low power consumption period, the monitoring device t (Unit: RMB / kWh, obtained through the communication module), controls the charging current I c =0.1C, charging time Where △SOC is the target state of charge change, and the calibration capacity C is calculated by the voltage-current integration method. a =∫I(t)dt, error <2%, generate a capacity report, record the remaining capacity and number of cycles.

[0025] Also includes: Based on the periodic change data of the feeding circuit current and power, the working status and health status evaluation results of the load equipment are generated; when the evaluation results indicate an abnormality, a load abnormality warning signal is generated and uploaded to the remote terminal via Ethernet or CAN bus.

[0026] Specifically, the monitoring device monitors the feed circuit current I t and power P t The data is Fourier transformed, and the input time series X(t) = [I t ,P t ], output frequency domain features Where k is the frequency index, N = 1024 (number of sampling points), extract the main frequency component, period f kThe random forest algorithm inputs the frequency domain feature F(k) and historical data, constructs 100 decision trees, with a maximum depth of D = 10 and feature weight W = [w I ,w P ], where w I =0.6 (current weight), w P = 0.4 (power weight), output load status (normal, overload, underload) and health status (failure probability P f ,>0.8 is abnormal). When the current fluctuation exceeds the threshold ΔI=±10%I n ,I n For the rated current, a load abnormality warning signal is generated, which includes the loop number and timestamp, encoded in SON format, and uploaded to the remote terminal via Ethernet or CAN bus to trigger operation and maintenance inspection.

[0027] Fault warning signals include at least one of the following: AC input overvoltage, undervoltage or phase loss; The DC bus or battery voltage is abnormal; The AC incoming line, busbar, battery or feeder switch trips; Rectifier module or lightning arrester failure; Abnormal battery charging or abnormal single cell voltage, internal resistance, and temperature.

[0028] Specifically, the monitoring device identifies the fault through the threshold detection algorithm, and the AC input over-voltage and under-voltage threshold V ac = ±15% V n ,V n =220V, phase loss through zero sequence voltage If V0>5V, it is determined to be phase loss; the DC bus voltage abnormal threshold ΔV dc = ±5% V dcn ,V dcn is the rated bus voltage; single cell voltage abnormality ΔV i = ±0.1V, abnormal internal resistance R i >5mΩ, abnormal temperature T i >80℃, float charge current abnormal I f >0.05°C. Switch tripping is detected through status signals, rectifier module faults are determined by temperature (>80°C) or output abnormality, and lightning arrester faults are detected through status feedback. Warning signals are encoded in JSON format ({"fault_type":"over_voltage","time":"2025-05-21T14:00:00","location":"bus_1"}) and uploaded via the communication module.

[0029] Example 2: Reference Figure 4In a second embodiment of the present invention, the present invention provides a digital intelligent DC power supply operation and maintenance system, comprising: a data acquisition module configured to obtain operating parameters of the DC power supply system through high-precision sensors, the operating parameters including battery cell voltage, current, internal resistance, and temperature, DC bus voltage and current, and feeder circuit voltage and current; a data processing module configured to perform digital filtering and machine learning analysis on operating parameters to generate battery state of charge, health status diagnostic results, and system status data; The control module is configured to dynamically adjust power supply output parameters according to system status data and generate fault diagnosis and processing suggestions; the communication module is configured to transmit operating parameters, system status data and processing suggestions through Ethernet or CAN bus.

[0030] Specifically, the data acquisition module acquires operating parameters, including battery cell voltage, current, internal resistance, and temperature, as well as DC bus voltage and current, and feeder circuit voltage and current, through a sensor interface at a 1kHz sampling rate. This data is stored in a local cache. The data processing module, running on an embedded processor, implements a Kalman filter algorithm (process noise covariance 0.01, measurement noise covariance 0.05, output noise <0.01%), a support vector machine algorithm (radial basis function kernel, kernel parameter 0.1, regularization parameter 1.0, SOC error <5%, SOH error <3%), and an ARIMA model (autoregressive order 2, differencing order 1, moving average order 1, predicted failure probability >80%) to generate SOC, SOH, and system status data. The control module adjusts the charger output voltage and current using a PWM signal (1ms cycle), supporting peak load shaving (reducing power by 10% during peak hours). The communication module utilizes an Ethernet switch or CAN controller, and data transmission complies with the DL / T459 protocol. The monitoring device integrates the above modules and coordinates their operation through an embedded operating system. It uses a feedforward neural network (3 hidden layers, 100 neurons per layer, Sigmoid activation function, learning rate 0.01, number of iterations 1000, and 5000 training data sets) to predict faults and support demand-side response functions.

[0031] The data processing module includes: a preprocessing unit configured to eliminate noise in the operating parameters by using a digital filtering technique to generate preprocessed data; an analysis unit configured to analyze the preprocessed data to generate SOC and SOH diagnosis results based on a machine learning algorithm; The prediction unit is configured to generate a fault warning signal based on historical and real-time operating parameters.

[0032] Specifically, the preprocessing unit of the data processing module runs the Kalman filter algorithm, inputs the original voltage and current data, sets the process noise covariance to 0.01, the measurement noise covariance to 0.05, and outputs preprocessed data with noise less than 0.01%. The analysis unit uses the support vector machine algorithm, inputs voltage, current, internal resistance, and temperature characteristics, and uses the radial basis function kernel (kernel parameter 0.1, regularization parameter 1.0) to generate SOC (error <5%) and SOH (error <3%) diagnostic results. The prediction unit uses the ARIMA model, inputs voltage, internal resistance, and temperature time series, sets the autoregressive order to 2, the difference order to 1, and the moving average order to 1, predicts the probability of failure (such as internal resistance abnormality >80%), outputs a warning signal, and stores it in the local database for operation and maintenance strategy call.

[0033] The control module includes: A power management unit is configured to dynamically adjust the output voltage and current of the charging module based on system status data to support peak shaving and valley filling; Desulfurization unit, configured to perform online desulfurization of the battery through low-current pulse charge and discharge technology, combining active and passive balancing to optimize battery performance; The operation and maintenance strategy unit is configured to generate fault handling suggestions based on the fault type knowledge graph.

[0034] Specifically, the power management unit adjusts the output of the charging module through the PI controller based on the system status data. The voltage adjustment range is ±10V, the current adjustment range is ±5A, and it supports peak shaving and valley filling (reducing power by 10% during peak hours). The desulfurization unit generates a 10Hz-100Hz pulse signal to control the charging module to output a 0.05C current, triggering the decomposition of lead sulfate crystals. Active balancing realizes power transfer through the DC-DC converter, and passive balancing dissipates excess power through parallel resistors. Based on the RDF knowledge graph, the operation and maintenance strategy unit queries fault characteristics (such as "internal resistance > 5mΩ"), matches processing suggestions (such as "replace monomer"), and outputs them in JSON format.

[0035] Example 3 The third embodiment of the present invention, based on the same inventive concept, proposes a digital intelligent DC power supply operation and maintenance device, comprising: processor; a memory storing instructions executable by a processor; Among them, when the processor executes the instructions, the steps of the above-mentioned digital DC power supply operation and maintenance method are implemented.

[0036] The processor is also configured to: Receive remote control commands and perform remote charging and discharging, parameter configuration or fault handling operations. Remote control commands comply with DL / T459, DL / T781, DL / T1074 and GB / T19826 power communication protocols; Operating parameters, system status data and fault warning signals are displayed through a visual interface.

[0037] Example 4 The fourth embodiment of the present invention is based on the same inventive concept. The present invention proposes a computer-readable storage medium, which stores a computer program. When the computer program is executed by a processor, the steps of the above-mentioned digital DC power supply operation and maintenance method are implemented.

[0038] It should be understood that various parts of the present invention can be implemented using hardware, software, firmware, or a combination thereof. In the above-described embodiments, multiple steps or methods can be implemented using software or firmware stored in a memory and executed by a suitable instruction execution system. For example, if implemented using hardware, as in another embodiment, any one of the following technologies known in the art or a combination thereof can be used: a discrete logic circuit having a logic gate circuit for implementing a logic function on a data signal, an application-specific integrated circuit having a suitable combination of logic gate circuits, a programmable gate array (PGA), a field programmable gate array (FPGA), etc.

[0039] Finally, it should be noted that the above is only a preferred embodiment of the present invention and is not intended to limit the present invention. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art can still modify the technical solutions described in the aforementioned embodiments or make equivalent substitutions for some of the technical features therein. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the scope of protection of the present invention.

Claims

1. A digitalized DC power supply operation and maintenance method, applied to a DC power supply system, with a monitoring device as the execution subject, is characterized by: The following steps are involved: Acquire the DC power system operating parameters collected by high-precision sensors, including battery cell voltage, current, internal resistance, temperature, DC bus voltage, current, and feeder circuit voltage and current; Performing digital filtering on the operating parameters to generate pre-processed data; Analyze the preprocessed data based on a machine learning algorithm to generate battery state of charge, health status diagnosis results and system status data; Dynamically adjust output parameters of the charging module based on the system status data, wherein the output parameters are used to support demand-side response and peak-shaving; Generate fault warning signals through fault prediction models based on historical and real-time operating parameters; Generate fault diagnosis and handling suggestions based on the fault type knowledge graph; The system status data, fault warning signals and processing suggestions are displayed through a visual interface and uploaded to a remote terminal via Ethernet or CAN bus.

2. The digital intelligent DC power supply operation and maintenance method according to claim 1, characterized in that: Also includes: The battery is desulfurized online through low-current pulse charge and discharge technology, and the optimized battery performance data is generated by combining active and passive balancing. During peak and valley periods of electricity consumption, quantitative charging and discharging are performed based on the pre-processed data to generate calibrated battery capacity data for online capacity verification.

3. The digital intelligent DC power supply operation and maintenance method according to claim 1, characterized in that: Also includes: Generate evaluation results of the working status and health status of the load equipment based on the periodic change data of the feeding circuit current and power; When the evaluation result indicates an abnormality, a load abnormality warning signal is generated and uploaded to a remote terminal via Ethernet or CAN bus.

4. The digital intelligent DC power supply operation and maintenance method according to claim 1, characterized in that: The fault warning signal includes at least one of the following: AC input overvoltage, undervoltage or phase loss; The DC bus or battery voltage is abnormal; The AC incoming line, busbar, battery or feeder switch trips; Rectifier module or lightning arrester failure; Abnormal battery charging or abnormal single cell voltage, internal resistance, and temperature.

5. Digital intelligent DC power supply operation and maintenance system, characterized by: The digital intelligent DC power supply operation and maintenance method according to any one of claims 1 to 4 comprises: A data acquisition module is configured to obtain operating parameters of the DC power supply system through high-precision sensors, wherein the operating parameters include battery cell voltage, current, internal resistance, temperature, DC bus voltage and current, and feed circuit voltage and current; a data processing module configured to perform digital filtering and machine learning analysis on the operating parameters to generate battery state of charge, health status diagnosis results and system status data; a control module configured to dynamically adjust power supply output parameters according to the system status data and generate fault diagnosis and treatment suggestions; A communication module is configured to transmit the operating parameters, system status data and processing suggestions via Ethernet or CAN bus.

6. The digital intelligent DC power supply operation and maintenance system according to claim 5, characterized in that: The data processing module includes: a preprocessing unit configured to eliminate noise in the operating parameters by using a digital filtering technique to generate preprocessed data; an analyzing unit configured to analyze the preprocessed data based on a machine learning algorithm to generate SOC and SOH diagnostic results; The prediction unit is configured to generate a fault warning signal based on historical and real-time operating parameters.

7. The digital intelligent DC power supply operation and maintenance system according to claim 5, characterized in that: The control module includes: a power management unit configured to dynamically adjust the output voltage and current of the charging module according to the system status data to support peak shaving and valley filling; Desulfurization unit, configured to perform online desulfurization of the battery through low-current pulse charge and discharge technology, combining active and passive balancing to optimize battery performance; The operation and maintenance strategy unit is configured to generate fault handling suggestions based on the fault type knowledge graph.

8. A digital intelligent DC power supply operation and maintenance device, characterized in that: include: processor; a memory storing instructions executable by the processor; When the processor executes the instruction, the steps of the digitalized DC power supply operation and maintenance method according to any one of claims 1 to 4 are implemented.

9. The digital intelligent DC power supply operation and maintenance device according to claim 8, characterized in that: The processor is further configured to: Receive remote control commands and perform remote charging and discharging, parameter configuration, or fault handling operations. The remote control commands comply with DL / T459, DL / T781, DL / T1074, and GB / T19826 power communication protocols; The operating parameters, system status data and fault warning signals are displayed through a visual interface.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the digitalized DC power supply operation and maintenance method according to any one of claims 1 to 4 are implemented.

Images