Energy storage power station integrated detection device and method thereof

Abstract

This invention provides an integrated testing device and method for energy storage power stations, belonging to the field of energy storage power station testing technology. The method includes transporting the equipment to the target power station, connecting the communication interface, automatically identifying the protocol and establishing a communication connection via an FPGA chip, collecting multi-dimensional power station equipment operation data, and using edge computing algorithms to analyze the equipment status in real time and provide real-time early warnings. This invention proposes an integrated design supporting single-person on-site operation. The programmable FPGA chip enables adaptive protocol identification, is compatible with various power station communication protocols, and possesses IP65 protection and multi-mode communication characteristics, adapting to complex power station environments and ensuring more stable operation. Furthermore, it utilizes refined collection and processing of multi-dimensional power station operation data, relying on heterogeneous computing and edge computing capabilities to achieve local real-time analysis, and employs dedicated analysis models for different devices to accurately assess equipment status from dimensions such as aging, thermal runaway, and consistency.

Description

Technical Field

[0001] This invention belongs to the field of energy storage power station testing technology, and particularly relates to an integrated testing device and method for energy storage power stations. Background Technology

[0002] Electrochemical energy storage power stations are costly to build, exhibit significant electrochemical characteristics, and possess abundant monitoring points, generating large volumes of data with extremely high analytical value. Currently, data acquisition in energy storage power stations primarily relies on traditional computer servers for data collection and analysis, lacking dedicated acquisition terminals and lightweight analysis tools. Traditional computer servers have stringent requirements regarding usage scenarios, spatial environments, and network conditions, and are often fixed in place, resulting in large size and weight, making them unsuitable for mobile environments. While server-based acquisition devices are suitable for long-term, uninterrupted daily early warning analysis of power stations, different power stations have varying early warning cycle requirements. Therefore, more flexible terminal devices are needed to meet the power station's need for continuous equipment monitoring and early warning analysis in its daily operating environment.

[0003] However, existing testing equipment for energy storage power stations has the following problems: Distributed architecture: Traditional detection relies on multiple independent devices (sensors, data loggers, servers), which are complex to deploy and have poor coordination.

[0004] Insufficient mobility: Fixed servers are difficult to adapt to the flexible testing needs of distributed power stations, and manual handling costs are high.

[0005] Poor hardware compatibility: Different power plants have different interface protocols and communication standards, making it difficult for existing equipment to adapt to multiple scenarios.

[0006] Data processing delay: Data needs to be transmitted back to the cloud for analysis, resulting in low real-time performance and dependence on network stability.

[0007] Insufficient algorithm-hardware coordination: Existing equipment lacks dedicated analysis algorithms tailored to the characteristics of energy storage power stations, and the algorithm execution is disconnected from the hardware data acquisition process, making real-time early warning impossible. Summary of the Invention

[0008] In view of the shortcomings of the prior art, the present invention provides an integrated testing device and method for energy storage power stations.

[0009] In a first aspect, the device includes a lightweight portable chassis, a multi-protocol data acquisition module, a heterogeneous computing main control module, a multi-mode communication module, an intelligent power management module, and an environmental adaptation module; The lightweight portable chassis features a lightweight aluminum alloy frame, built-in silicone pads and a honeycomb structure shock-absorbing layer, and is equipped with a folding pull rod and wheels to support single-person movement. The multi-protocol data acquisition module includes a multi-protocol interface array, a signal conditioning unit, a sensor expansion slot, and a programmable FPGA chip; wherein, the multi-protocol interface array specifically includes an integrated industrial communication interface array, which supports hot-swapping and is directly connected to the programmable FPGA chip; the programmable FPGA chip has a built-in protocol feature library and a machine learning classification model for adaptive protocol identification; The heterogeneous computing main control module includes a domestically produced multi-core CPU, a GPU acceleration unit, a high-speed cache layer, and a persistent storage layer; The multi-mode communication module includes 5G / 4G, Wi-Fi 6, LoRaWAN wireless transmission units and a local networking unit; The intelligent power management module includes a wide voltage input unit, a high-density battery pack, an intelligent power distribution unit, and a breakpoint resume mechanism. The environmental adaptation module includes a heat dissipation system, an IP65 protection unit, and a human-machine interaction unit.

[0010] Secondly, based on the aforementioned device, the method includes, After transporting the equipment to the target power station and connecting the communication interface, the FPGA chip automatically identifies the protocol and establishes a communication connection. Collect multi-dimensional power plant equipment operation data and use edge computing algorithms to analyze equipment status in real time and provide real-time early warnings.

[0011] Furthermore, after connecting to the communication interface, the FPGA chip automatically identifies the protocol and establishes a communication connection, specifically including: Transport the equipment to the target energy storage power station site; select dedicated cables according to the interface type of the power station equipment, connect the communication interface of the power station EMS to the multi-protocol interface array, and connect external sensors to the sensor expansion slot according to the testing requirements; After establishing a communication connection, a self-test is performed, the initial power consumption of each module is dynamically allocated, and the device enters the working mode; the programmable FPGA chip starts the protocol identification process, initializes the internal protocol feature library and machine learning classification model, and prepares to receive handshake signals at the interface. The handshake signal emitted by the power station's EMS communication interface is acquired by a multi-protocol interface array, and the key features of the handshake signal are extracted. The key features of the extracted handshake signals are matched one by one with the internally stored protocol feature library, and the corresponding driver configuration file is loaded. After the driver configuration file is loaded, a data acquisition request command is sent to the power station EMS, and the response signal from the power station EMS is received to verify the stability and effectiveness of the communication connection.

[0012] Furthermore, the step of matching the key features of the extracted handshake signals one by one with the internally stored protocol feature library and loading the corresponding driver configuration file specifically includes: If a match is successful, the programmable FPGA chip automatically identifies the corresponding protocol type, calls and loads the corresponding driver configuration file, and completes protocol parsing and communication parameter configuration. If a match fails, the programmable FPGA chip starts a machine learning classification model, inputting the collected unknown protocol features into the model for classification and recognition.

[0013] Furthermore, the verification of the stability and validity of the communication connection specifically includes, If a normal response signal is received, it indicates that the communication connection is successful and the device enters the data acquisition preparation state. If no response signal is received or the response signal is abnormal, the device will re-execute the protocol identification process and repeat it a certain number of times. If effective communication still cannot be established, an alarm message will be sent.

[0014] Furthermore, the collection of multi-dimensional power plant equipment operation data and the use of edge computing algorithms to analyze equipment status in real time and provide real-time early warnings specifically include: After the communication connection is successfully established, configure the relevant data acquisition parameters according to the preset detection scheme; Based on the configured data acquisition parameters, collect multi-dimensional raw signals from the energy storage power station; The collected analog signals are processed to obtain the mathematical signals of the energy storage power station, and then classified and stored according to the time series. Edge computing algorithms are used to perform real-time analysis and processing of mathematical signals from energy storage power stations to obtain analysis results; If the analysis results contain early warning information, the early warning information will be classified and organized, the early warning level will be marked, and the corresponding early warning response mechanism will be triggered.

[0015] Furthermore, the refined processing of the acquired analog signals specifically includes isolating and amplifying the acquired analog signals, filtering them, and performing AD conversion.

[0016] Furthermore, the use of edge computing algorithms to perform real-time analysis and processing of mathematical signals from the energy storage power station specifically includes: For transformers and converters in energy storage power stations, the aging status of transformers and converters in energy storage power stations is assessed by analyzing temperature trend data at different time dimensions. An LSTM-based time-series prediction model is used to predict the probability of thermal runaway risk in battery cells by analyzing historical operating data of the cells. The consistency of the battery cluster is assessed by analyzing the voltage and SOC data of each cell within the cluster.

[0017] Furthermore, the assessment of the aging status of the transformers and converters in the energy storage power station by analyzing temperature trend data across different time dimensions specifically includes: Extract daily, monthly, and yearly temperature data sequences from transformers and converters to generate temperature trend curves; Based on the temperature trend curve, calculate the temperature change rate and the heat dissipation coefficient λ. Based on the calculated temperature change rate and heat dissipation coefficient, calculate the equipment aging coefficient; The calculated aging coefficient is compared with a preset threshold.

[0018] Furthermore, the aforementioned LSTM-based time-series prediction model, through analysis of historical operating data of the battery cells, specifically includes: Extract key parameter data within a certain time period to form a time series data sequence; An LSTM time series prediction model is constructed and trained and optimized using adaptive moment estimation to minimize the prediction error; The probability of thermal runaway risk is calculated using the trained LSTM time series prediction model. Early warning judgments are made based on the calculated probability P of thermal runaway risk.

[0019] Furthermore, the method of evaluating the consistency of the battery cluster by analyzing the voltage and SOC data of each cell within the cluster specifically includes: Extract daily and monthly voltage trend data and SOC data of battery clusters to form voltage and SOC sequences for each cell; Based on the voltage sequence and SOC sequence, the standard deviation of voltage and the deviation of SOC of cells within the battery cluster are calculated. The consistency coefficient C of the battery cluster is calculated using the consistency coefficient formula. The calculated consistency coefficient C is compared with the preset threshold. Comparisons are made to establish early warning judgments.

[0020] Compared with the prior art, the present invention has the following advantages: 1. This invention proposes an integrated design that supports single-person on-site operation. The programmable FPGA chip enables adaptive protocol recognition, is compatible with various power plant communication protocols, and has features such as IP65 protection and multi-mode communication. It can adapt to the complex on-site environment of power plants and operate more stably.

[0021] 2. Utilize refined collection and processing of multi-dimensional power plant operation data, and rely on heterogeneous computing and edge computing capabilities to achieve local real-time analysis. Employ dedicated analysis models for different equipment to accurately assess equipment status from dimensions such as aging, thermal runaway, and consistency.

[0022] 3. The system provides graded early warnings based on test results and triggers corresponding response mechanisms. Intelligent power management can dynamically allocate power consumption and enable breakpoint resume transmission. The fault-tolerant design of the communication and self-test links further ensures the smoothness of the testing process, providing comprehensive and reliable data support and risk prevention for power plant operation and maintenance.

[0023] Other features and advantages of the invention will be set forth in the description which follows, and will be apparent in part from the description, or may be learned by practicing the invention. The objects and other advantages of the invention may be realized and obtained by means of the structures pointed out in the description, claims and drawings. Attached Figure Description

[0024] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0025] Figure 1 A schematic diagram of an integrated testing equipment module for an energy storage power station according to the present invention is shown.

[0026] Figure 2 A schematic diagram of the integrated testing method for energy storage power stations according to the present invention is shown. Detailed Implementation

[0027] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0028] like Figure 1 As shown, this invention provides an integrated testing device for energy storage power stations, comprising: a lightweight portable chassis, a multi-protocol data acquisition module, a heterogeneous computing main control module, a multi-mode communication module, an intelligent power management module, and an environmental adaptation module. Each module is isolated by an electromagnetic shielding partition and achieves data interaction through a high-speed PCIe 4.0 bus (main control module and storage module), an SPI bus (main control module and power management module), an LVDS bus (main control module and human-machine interface unit), and a gigabit Ethernet (acquisition module and communication module).

[0029] In this embodiment, the lightweight portable chassis adopts a lightweight aluminum alloy frame, with built-in silicone pads and a honeycomb structure shock-absorbing layer. The dimensions are ≤600mm×400mm×200mm, the weight is ≤10kg, and it is equipped with a folding pull rod and wheels to support single-person movement.

[0030] The multi-protocol data acquisition module includes a multi-protocol interface array, a signal conditioning unit, a sensor expansion slot, and a programmable FPGA chip; among which, The signal conditioning unit includes a 24-bit high-precision ADC, an isolation amplifier with a withstand voltage of ≥2500V, and an 8th-order Butterworth low-pass filter; the FPGA chip has a built-in protocol feature library and machine learning classification model for adaptive protocol identification. Optionally, the multi-protocol data acquisition module integrates industrial communication interface arrays such as RS-485 / 232, CAN, and Modbus-TCP, supports hot-swapping, and the interface array is directly connected to the programmable FPGA chip; wherein, the FPGA chip realizes protocol parsing and conversion through a preset configuration file.

[0031] The heterogeneous computing main control module includes a domestically produced multi-core CPU, a GPU acceleration unit, a high-speed cache layer, and a persistent storage layer.

[0032] The multi-mode communication module includes 5G / 4G, Wi-Fi 6, LoRaWAN wireless transmission units and a local networking unit; The intelligent power management module includes a wide-voltage input unit, a high-density battery pack, an intelligent power distribution unit, and a breakpoint resume mechanism; The environmental adaptation module includes a heat dissipation system, an IP65 protection unit, and a human-machine interaction unit.

[0033] like Figure 2 As shown, this invention proposes an integrated testing method for energy storage power stations. It achieves adaptive connection with different power station equipment through the multi-protocol interface array of a portable device and the protocol recognition capability of an FPGA chip; utilizes the device's signal conditioning unit and sensor expansion slots to complete the acquisition and preprocessing of multi-dimensional raw signals; achieves real-time edge computing by executing dedicated analysis algorithms for key equipment in the energy storage power station in parallel through a CPU+GPU heterogeneous computing unit; finally, it generates a detailed testing report and ensures the integrity of data transmission and storage through a multi-mode communication module and a breakpoint resume mechanism. The steps include... S1. Transport the portable testing equipment to the target power station. After connecting the communication interface, the equipment automatically identifies the protocol and establishes a communication connection through the FPGA chip.

[0034] S1.1 Transport the portable testing equipment to the target energy storage power station site by a single person; and select a dedicated cable according to the interface type of the power station equipment to connect the communication interface of the power station EMS (Energy Management System) to the multi-protocol interface array of the equipment. At the same time, according to the testing requirements, connect external sensors such as voltage / current transformers, temperature sensors, and vibration sensors to the sensor expansion slot of the equipment.

[0035] Alternatively, if the energy storage power station supports wired networking, the power station equipment can be connected to the local networking unit (RJ45 interface) of the equipment via a gigabit network cable to improve communication stability; if the energy storage power station does not support wired networking, a wireless communication connection can be established through the multi-mode wireless transmission unit of the equipment.

[0036] S1.2 After establishing a communication connection, a self-test is performed, the initial power consumption of each module is dynamically allocated, and the device enters the working mode; the FPGA chip starts the protocol identification process, initializes the internal protocol feature library and machine learning classification model, and prepares to receive the handshake signal at the interface; at the same time, the touch screen of the human-machine interaction unit lights up and displays the prompt message "Protocol identification in progress, please wait", and the status indicator light flashes blue.

[0037] S1.3 Acquire the handshake signal sent by the power station EMS communication interface through a multi-protocol interface array, and extract the key features of the handshake signal.

[0038] In this embodiment, during the acquisition process, the FPGA chip ensures the accuracy of signal feature acquisition through a high-speed sampling circuit, and the sampling frequency is automatically adjusted according to the interface type to ensure that no key signal features are missed.

[0039] S1.4. Match the key features of the extracted handshake signal one by one with the internally stored protocol feature library (which includes feature parameters of commonly used power plant protocols such as RS-485 / 232, CAN, and Modbus-TCP) and load the corresponding driver configuration file.

[0040] In this embodiment, if the match is successful, the FPGA chip automatically identifies the corresponding protocol type, calls and loads the driver configuration file corresponding to the protocol from the device storage module, and completes the protocol parsing and communication parameter configuration. If a match fails, the FPGA chip initiates a machine learning classification model (based on the random forest algorithm), inputting the collected unknown protocol features into the model for classification and identification. The model's training set contains features of various power plant protocols, giving it a strong ability to identify unknown protocols. Through multi-dimensional analysis of the features, a temporary driver configuration file is generated and loaded to achieve communication adaptation with devices using unknown protocols.

[0041] S1.5 After the driver configuration file is loaded, a data acquisition request command is sent to the power station EMS, and the response signal from the power station EMS is received to verify the stability and effectiveness of the communication connection.

[0042] In this embodiment, if a normal response signal is received, it indicates that the communication connection is successful and the device enters the data acquisition preparation state; if no response signal is received or the response signal is abnormal, the device will re-execute the protocol identification process, repeating it up to 3 times. If effective communication still cannot be established, an alarm message will be sent.

[0043] S2. Collect multi-dimensional power plant equipment operation data and use edge computing algorithms to analyze equipment status in real time and provide real-time early warning.

[0044] S2.1 After the communication connection is successfully established, configure the relevant data acquisition parameters (such as acquisition period, data type, sampling frequency and storage strategy) according to the preset detection scheme.

[0045] S2.2. Collect multi-dimensional raw signals from the energy storage power station according to the configured data acquisition parameters.

[0046] S2.3. The collected analog signals are refined to obtain the mathematical signals of the energy storage power station, and then classified and stored according to the time series (day, month, year).

[0047] In this embodiment, the present invention performs fine processing on the acquired analog signals, including: 1. Isolation Amplification: The signal is isolated and amplified by the isolation amplifier to avoid damage to the internal circuit of the equipment by the high voltage signal. At the same time, the weak signal is amplified to a suitable acquisition range. The amplification gain can be automatically adjusted according to the signal strength, up to 100 times.

[0048] 2. Filtering Process: An 8th-order Butterworth low-pass filter is used for anti-aliasing filtering of the signal. The cutoff frequency can be adjusted via FPGA programming (adjustment range: 10Hz~100kHz), effectively filtering out high-frequency noise and interference signals in the signal. The filter circuit consists of an OPA227 operational amplifier and precision resistors and capacitors to ensure the stability and reliability of the filtering effect.

[0049] 3. AD Conversion: The conditioned analog signal is converted into a digital signal using a 24-bit high-precision ADC. Differential sampling is employed during the conversion process to improve anti-interference capability and conversion accuracy. The digital signal is connected to the FPGA chip via the SPI bus, and then transmitted from the FPGA chip to the main control module.

[0050] S2.4. Using edge computing algorithms, the mathematical signals of the energy storage power station are analyzed and processed in real time to obtain the analysis results.

[0051] In this embodiment, the edge computing algorithm used in this invention mainly includes: 1. Equipment Aging Assessment Algorithm: Primarily targeting transformers and converters in energy storage power stations, this algorithm analyzes temperature trend data across different time dimensions (daily, monthly, and yearly) to assess the aging status of the equipment. The steps include: a. Data input: Extract daily, monthly, and yearly temperature data sequences of transformers and converters from the storage module to form temperature trend curves.

[0052] b. Calculation of characteristic parameters: Based on the temperature trend curve, calculate the temperature change rate k (unit: ℃ / day or ℃ / month) and the heat dissipation coefficient λ (unit: W / ℃); where the temperature change rate k reflects the rate of change of equipment temperature over time and is calculated by a linear fitting algorithm; the heat dissipation coefficient λ reflects the heat dissipation capacity of the equipment and is calculated based on the heat conduction equation and equipment structural parameters, i.e., λ=Q / ΔT, where Q is the heat dissipation power of the equipment and ΔT is the temperature difference between the equipment and the environment.

[0053] c. Aging coefficient calculation: Based on the calculated temperature change rate and heat dissipation coefficient, the equipment aging coefficient S is calculated, and its formula is expressed as follows: S = k × (1 / λ) Among them, the aging coefficient S comprehensively reflects the impact of the equipment temperature change rate and heat dissipation capacity on equipment aging. The larger the S value, the more serious the degree of equipment aging.

[0054] d. Early warning judgment: Compare the calculated aging coefficient S with the preset threshold. Compare. Threshold Custom settings can be configured based on equipment type, service life, and power plant operation and maintenance requirements. Default values ​​are available. =1.

[0055] Optionally, when When this occurs, an aging warning is triggered, generating the warning message "The transformer / converter is at a high level of aging; it is recommended to shorten the inspection cycle and perform targeted maintenance." When this time is reached, it indicates that the equipment is in normal aging condition and no warning information is generated.

[0056] 2. Cell Thermal Runaway Prediction Algorithm: An LSTM-based time-series prediction model is used to predict the probability of thermal runaway risk by analyzing historical operating data of the battery cell. The steps include: a. Data Input: Extract the cell voltage difference ΔV (unit: mV), internal resistance change rate ΔR (unit: % / day), and temperature gradient from the storage module over a certain period (e.g., 30 days). Key parameter data such as T (unit: ℃ / cm) are used to form a time series data sequence.

[0057] b. Model Construction and Optimization: An LSTM time series prediction model is constructed, which includes an input layer, three hidden layers (128 neurons per layer) and one output layer. During training, the adaptive moment estimation (Adam) optimization algorithm is used to minimize the prediction error (cross-entropy loss function).

[0058] c. Calculation of thermal runaway risk probability: Calculate the thermal runaway risk probability using the trained LSTM time series prediction model.

[0059] In this embodiment, the present invention converts time-series data sequences into feature vectors that the model can process; and uses the ReLU activation function to enhance the nonlinear fitting ability of the model, and effectively captures long-term dependencies in time-series data through gating mechanisms (input gate, forget gate, output gate); and uses the Sigmoid activation function to map the prediction results to the [0,1] interval and output the probability P of thermal runaway risk.

[0060] Optionally, during device operation, the model parameters can be updated via a cloud platform to continuously optimize prediction accuracy.

[0061] d. Early warning judgment: Based on the calculated probability P of thermal runaway risk, an early warning judgment is made. Optionally, when the probability of thermal runaway risk P>0.7, a thermal runaway warning is triggered, and a warning message "The risk of thermal runaway of the battery cell is high, please stop the machine immediately for inspection" is generated; when P≤0.7, it indicates that the battery cell is operating normally and no warning message is generated.

[0062] 3. Battery Cluster Consistency Assessment Algorithm: By analyzing the voltage and SOC data of each cell within the battery cluster, the consistency status of the battery cluster is assessed. The steps include: a. Data input: Extract daily and monthly voltage trend data and SOC data of battery clusters from the storage module to form voltage sequence and SOC sequence of each cell.

[0063] b. Characteristic parameter calculation: Calculate the voltage standard deviation σ and SOC deviation ΔSOC of the cells within the battery cluster; where the voltage standard deviation σ is calculated using the overall standard deviation formula, which is expressed as follows:

[0064] Where n is the number of cells in the battery cluster that are included in the calculation; Let be the measured voltage value of the i-th cell; The average voltage value of the cells involved in the calculation within the battery cluster.

[0065] Optionally, the voltage standard deviation σ reflects the dispersion of the voltage of each cell. The larger the value of σ, the worse the voltage consistency. The SOC deviation value ΔSOC is the difference between the maximum and minimum SOC values ​​of each cell in the battery cluster. The larger the value of ΔSOC, the worse the SOC consistency.

[0066] c. Consistency Coefficient Calculation: The consistency coefficient C of the battery cluster is calculated using the consistency coefficient formula, where the consistency coefficient formula is expressed as: C = 1 / (σ×ΔSOC) Among them, the consistency coefficient C is positively correlated with the consistency of the battery cluster. The larger the C value, the better the consistency of the battery cluster; the smaller the C value, the worse the consistency of the battery cluster.

[0067] d. Early warning judgment: Compare the calculated consistency coefficient C with the preset threshold. Comparison; where the threshold Custom settings can be configured based on battery cluster type, service life, and power plant operation and maintenance requirements. Default values ​​are available. =0.5. When When this occurs, a consistency warning is triggered, generating the warning message "Battery cluster consistency is poor; it is recommended to perform cell balancing or replace abnormal cells." When this occurs, it indicates that the battery cluster consistency is normal and no warning information is generated.

[0068] S2.5 If the analysis results contain early warning information, the early warning information shall be classified and organized, the early warning level shall be marked, and the corresponding early warning response mechanism shall be triggered; among them, the early warning level is divided into general level, important level, and emergency level.

[0069] The foregoing description and accompanying drawings fully illustrate embodiments of the invention to enable those skilled in the art to practice them. Other embodiments may include structural and other changes. The embodiments represent only possible variations. Individual components and functions are optional unless explicitly required, and the order of operation may vary. Some portions and features of some embodiments may be included or substituted for portions and features of other embodiments. Embodiments of the invention are not limited to the structures described above and shown in the accompanying drawings, and various modifications and changes may be made without departing from their scope. The scope of the invention is limited only by the appended claims.

Claims

1. An integrated detection device for energy storage power stations, characterized by, The device includes a lightweight portable chassis, a multi-protocol data acquisition module, a heterogeneous computing main control module, a multi-mode communication module, an intelligent power management module, and an environmental adaptation module. The lightweight portable chassis features a lightweight aluminum alloy frame, built-in silicone pads and a honeycomb structure shock-absorbing layer, and is equipped with a folding pull rod and wheels to support single-person movement. The multi-protocol data acquisition module includes a multi-protocol interface array, a signal conditioning unit, a sensor expansion slot, and a programmable FPGA chip; wherein, the multi-protocol interface array specifically includes an integrated industrial communication interface array, which supports hot-swapping and is directly connected to the programmable FPGA chip; the programmable FPGA chip has a built-in protocol feature library and a machine learning classification model for adaptive protocol identification; The heterogeneous computing main control module includes a domestically produced multi-core CPU, a GPU acceleration unit, a high-speed cache layer, and a persistent storage layer; The multi-mode communication module includes 5G / 4G, Wi-Fi 6, LoRaWAN wireless transmission units and a local networking unit; The intelligent power management module includes a wide voltage input unit, a high-density battery pack, an intelligent power distribution unit, and a breakpoint resume mechanism. The environmental adaptation module includes a heat dissipation system, an IP65 protection unit, and a human-machine interaction unit.

2. A method for integrated detection of an energy storage power plant, characterized in that Based on the device of claim 1, the method includes, After transporting the equipment to the target power station and connecting the communication interface, the FPGA chip automatically identifies the protocol and establishes a communication connection. Collect multi-dimensional power plant equipment operation data and use edge computing algorithms to analyze equipment status in real time and provide real-time early warnings.

3. The method of claim 2, wherein the method further comprises: After connecting to the communication interface, the FPGA chip automatically identifies the protocol and establishes a communication connection, specifically including: Transport the equipment to the target energy storage power station site; select dedicated cables according to the interface type of the power station equipment, connect the communication interface of the power station EMS to the multi-protocol interface array, and connect external sensors to the sensor expansion slot according to the testing requirements; After establishing a communication connection, a self-test is performed, the initial power consumption of each module is dynamically allocated, and the device enters the working mode; the programmable FPGA chip starts the protocol identification process, initializes the internal protocol feature library and machine learning classification model, and prepares to receive handshake signals at the interface. The handshake signal emitted by the power station's EMS communication interface is acquired by a multi-protocol interface array, and the key features of the handshake signal are extracted. The key features of the extracted handshake signals are matched one by one with the internally stored protocol feature library, and the corresponding driver configuration file is loaded. After the driver configuration file is loaded, a data acquisition request command is sent to the power station EMS, and the response signal from the power station EMS is received to verify the stability and effectiveness of the communication connection.

4. The integrated testing method for energy storage power stations according to claim 3, characterized in that, The process of matching the key features of the extracted handshake signals one by one with the internally stored protocol feature library and loading the corresponding driver configuration file specifically includes: If a match is successful, the programmable FPGA chip automatically identifies the corresponding protocol type, calls and loads the corresponding driver configuration file, and completes protocol parsing and communication parameter configuration. If a match fails, the programmable FPGA chip starts a machine learning classification model, inputting the collected unknown protocol features into the model for classification and recognition.

5. The integrated testing method for energy storage power stations according to claim 3, characterized in that, The verification of the stability and validity of the communication connection specifically includes, If a normal response signal is received, it indicates that the communication connection is successful and the device enters the data acquisition preparation state. If no response signal is received or the response signal is abnormal, the device will re-execute the protocol identification process and repeat it a certain number of times. If effective communication still cannot be established, an alarm message will be sent.

6. The integrated detection method of the energy storage power station according to claim 2, characterized in that, The process of collecting multi-dimensional power plant equipment operation data and using edge computing algorithms to analyze equipment status in real time and provide real-time early warnings specifically includes: After the communication connection is successfully established, configure the relevant data acquisition parameters according to the preset detection scheme; Based on the configured data acquisition parameters, collect multi-dimensional raw signals from the energy storage power station; The collected analog signals are processed in a refined manner to obtain the mathematical signals of the energy storage power station, and then classified and stored according to the time series. Edge computing algorithms are used to perform real-time analysis and processing of mathematical signals from energy storage power stations to obtain analysis results; If the analysis results contain early warning information, the early warning information will be classified and organized, the early warning level will be marked, and the corresponding early warning response mechanism will be triggered.

7. The method of claim 6, wherein the method further comprises: The fine processing of the acquired analog signals specifically includes isolating and amplifying the acquired analog signals, filtering them, and performing AD conversion.

8. The integrated detection method of the energy storage power station according to claim 6, characterized in that, The method of using edge computing algorithms to perform real-time analysis and processing of mathematical signals from energy storage power stations specifically includes: For transformers and converters in energy storage power stations, the aging status of transformers and converters in energy storage power stations is assessed by analyzing temperature trend data at different time dimensions. An LSTM-based time-series prediction model is used to predict the probability of thermal runaway risk in battery cells by analyzing historical operating data of the cells. The consistency of the battery cluster is assessed by analyzing the voltage and SOC data of each cell within the cluster.

9. The integrated testing method for energy storage power stations according to claim 8, characterized in that, The method involves analyzing temperature trend data across different time dimensions to assess the aging status of the transformers and converters in the energy storage power station. Specifically, this includes... Extract daily, monthly, and yearly temperature data sequences from transformers and converters to generate temperature trend curves; Based on the temperature trend curve, calculate the temperature change rate and the heat dissipation coefficient λ. Based on the calculated temperature change rate and heat dissipation coefficient, calculate the equipment aging coefficient; The calculated aging coefficient is compared with a preset threshold.

10. The integrated testing method for energy storage power stations according to claim 8, characterized in that, The aforementioned LSTM-based time-series prediction model analyzes historical operating data of the battery cells, specifically including: Extract key parameter data within a certain time period to form a time series data sequence; An LSTM time series prediction model is constructed and trained and optimized using adaptive moment estimation to minimize the prediction error; The probability of thermal runaway risk is calculated using the trained LSTM time series prediction model. Early warning judgments are made based on the calculated probability P of thermal runaway risk.

11. The integrated testing method for energy storage power stations according to claim 8, characterized in that, The method of evaluating the consistency of the battery cluster by analyzing the voltage and SOC data of each cell within the cluster includes, specifically, Extract daily and monthly voltage trend data and SOC data of battery clusters to form voltage and SOC sequences for each cell; Based on the voltage sequence and SOC sequence, the standard deviation of voltage and the deviation of SOC of cells within the battery cluster are calculated. The consistency coefficient C of the battery cluster is calculated using the consistency coefficient formula. The calculated consistency coefficient C is compared with the preset threshold. Comparisons are made to establish early warning judgments.

Images