Charging cabin controller and battery safety linkage method and charging equipment

Abstract

The embodiment of the present disclosure provides a kind of charging cabin controller and the safety linkage method of battery and charging equipment, charging cabin controller controls the charging process of battery by charging device, battery has wireless communication function, safety linkage method includes: charging cabin controller generates connection quality grade by obtaining the plugging state between battery and charging device;When responding to connection quality grade greater than preset quality grade threshold, execute pre-charging self-checking stage, by the interaction between charging cabin controller and battery, the state of battery is determined by charging cabin controller;When responding to the state of battery is normal state, enter pre-charging stage, the current state information from battery is obtained by electric cabin controller, and current state information is sent to charging device;When responding to the confirmation result of charging cabin controller and charging device for current state information all satisfy set requirement, enter charging stage.The safety in charging process can be improved by using the above scheme.

Description

Safety linkage method between charging compartment controller and battery and charging equipment Technical Field

[0001] The embodiments disclosed in this invention relate to the field of charging control technology, and in particular to a method and charging device for safe linkage between a charging compartment controller and a battery. Background Technology

[0002] With the rapid development of battery technology, rechargeable batteries are widely used in energy storage, transportation, and consumer electronics. However, safety issues during battery charging are becoming increasingly prominent, such as increased contact resistance and localized overheating due to poor connector mating, or overcharging and thermal runaway caused by abnormal internal battery conditions (such as short circuits or aging).

[0003] Traditional charging control methods often rely on simple voltage or current threshold detection, lacking a comprehensive assessment of connection quality and battery status, making it difficult to identify potential risks in real time.

[0004] In existing solutions, charging control is typically performed independently by the charging device or the battery management system. However, insufficient coordination between these parties can lead to information delays or misjudgments. Against this backdrop, how to comprehensively ensure safety during the charging process through an intelligent safety linkage method has become an urgent technical problem to be solved. Summary of the Invention

[0005] In view of this, the present disclosure provides a method for safe linkage between a charging compartment controller and a battery, as well as a charging device, which can improve the safety of the charging process.

[0006] In a first aspect, embodiments of this disclosure provide a safety linkage method between a charging compartment controller and a battery. The charging compartment controller controls the charging process of the battery through a charging device. The battery has wireless communication capabilities. The safety linkage method includes:

[0007] The charging compartment controller generates a connection quality level by acquiring the connection status between the battery and the charging device.

[0008] When the connection quality level is greater than a preset quality level threshold, a pre-charging self-test phase is executed. Through the interaction between the charging case controller and the battery, the state of the battery is determined by the charging case controller.

[0009] In response to the battery being in a normal state, the pre-charging phase begins. The charging case controller obtains the current state information from the battery and sends the current state information to the charging device.

[0010] In response to the confirmation results of the current status information by the charging compartment controller and the charging device both meeting the set requirements, the charging phase begins, and the charging compartment controller controls the charging process of the battery; wherein, when the charging compartment controller detects an abnormality in the charging process, it outputs a stop command to the charging device to stop the current charging process.

[0011] Optionally, the charging case controller generates a connection quality level by acquiring the connection status between the battery and the charging device, including:

[0012] During the process of connecting the charging device and the battery through the connector, the charging compartment controller synchronously collects the spatiotemporal sequence data of the connection point to construct a multidimensional spatiotemporal observation dataset;

[0013] The charging compartment controller establishes a parameterized forward model describing the multiphysics coupling relationship of the plug-in point, and constructs a loss function based on the difference between the predicted output of the forward model and the multidimensional spatiotemporal observation dataset. Combining the attribute information of the connector, a gradient optimization algorithm is used to inversely solve for the intrinsic parameter set of the plug-in point.

[0014] Based on the intrinsic parameter set, the charging compartment controller establishes a causal directed acyclic graph with the physical parameters of the connector as nodes and the physical mechanism as edges, and performs virtual intervention analysis on the causal directed acyclic graph to calculate the causal effect strength of each intrinsic parameter on the connection quality and identify key causal parameters.

[0015] The charging case controller determines the connection quality level based on the values ​​and uncertainties of the key causal parameters.

[0016] Optionally, the safety linkage method satisfies one or more of the following:

[0017] The spatiotemporal sequence data includes: current, voltage, temperature, and at least two of the vibrations; wherein the voltage is the voltage division value on the CC2 signal line;

[0018] The intrinsic parameter set includes: contact resistance, actual contact area, and locking force;

[0019] The physical parameters of the connector include one or more of the following: loop impedance, insertion displacement, and contact pressure;

[0020] The step of determining the connection quality level based on the values ​​and uncertainties of the key causal parameters includes: using a Bayesian inference framework, propagating the values ​​and uncertainties of the key causal parameters as random variables, and calculating the confidence interval of the connection quality level through the updated posterior distribution.

[0021] Optionally, in response to the connection quality level being greater than a preset quality level threshold, a pre-charging self-test phase is performed, wherein the charging case controller determines the state of the battery through interaction between the charging case controller and the battery, including:

[0022] The charging compartment controller applies a wide-spectrum pulse voltage excitation to the battery, and simultaneously collects the time-domain response data throughout the process.

[0023] Based on the full-process time-domain response data, the charging compartment controller constructs a joint feature vector describing the static and dynamic characteristics of the battery;

[0024] The joint feature vector is fused with the long-term time-series operating feature vector of the battery under natural operating conditions to construct a multimodal feature tensor;

[0025] The multimodal feature tensor is input into a spatiotemporal graph neural network model based on an attention mechanism. The spatiotemporal graph neural network model uses battery cells as nodes and physical connections as edges. It captures spatial dependencies through graph convolution and analyzes feature evolution trends based on temporal convolution to generate temporal evolution results.

[0026] The state of the battery is determined by the charging compartment controller based on the spatial dependence and the temporal evolution results.

[0027] Optionally, the step of constructing a joint feature vector describing the static and dynamic characteristics of the battery by the charging compartment controller based on the full-process time-domain response data includes: performing modal decomposition on the time-domain response data by the charging compartment controller to decouple the signal into a first component characterizing the forced response and a second component characterizing the inherent oscillation characteristics; performing frequency domain analysis on the first component to determine the impedance spectrum of the battery and extracting the amplitude frequency response and phase frequency response to form a frequency response feature set; performing modal parameter identification based on the second component to extract the modal parameters of the battery to form a modal parameter set; determining the set of interaction terms between the frequency response feature set and the modal parameter set, and selecting interaction terms with significance higher than a threshold from the set of interaction terms; and using the interaction terms to structurally concatenate the frequency response feature set and the modal parameter set to form the joint feature vector.

[0028] The step of fusing the joint feature vector with the long-term time-series operating feature vector of the battery under natural operating conditions to construct a multimodal feature tensor includes: aligning the joint feature vector with the long-term time-series operating feature vector, projecting the aligned feature vector onto a unified feature space through an encoder network, and then obtaining the multimodal feature tensor through tensor concatenation.

[0029] Optionally, during the process of determining the state of the battery through the interaction between the battery and the charging compartment controller, the battery also establishes communication with the charging device, and the charging device requests the battery to send its own state information; in response to not receiving a reply signal from the battery within a preset time, the charging process is terminated.

[0030] Optionally, in response to the confirmation results from both the charging compartment controller and the charging device regarding the current state information meeting set requirements, the charging phase begins, and the charging compartment controller controls the charging process of the battery, including:

[0031] The charging device evaluates the battery’s required charging voltage and current against its own output capability parameters to determine whether the charging conditions are met.

[0032] In response to the fulfillment of the charging conditions, a two-way communication confirmation is performed between the charging device and the battery to verify the integrity and rationality of the received information;

[0033] The charging compartment controller independently assesses charging safety based on the current status information.

[0034] When the charging conditions are met, bidirectional communication is successfully confirmed, and the charging compartment controller does not detect any abnormality, charging is authorized to begin.

[0035] In response to any exception, stop the current charging process.

[0036] Optionally, before generating the connection quality level, the method further includes:

[0037] The charging compartment controller generates an environmental safety score based on the collected environmental parameters inside the charging compartment in the coal mine, and when the environmental safety score is greater than a preset safety score threshold, it executes the generation of connection quality level.

[0038] The step of generating an environmental safety score based on collected environmental parameters within the charging compartment in the coal mine includes: collecting multi-source environmental parameters such as combustible gas concentration, compartment temperature, humidity, and ventilation status; comparing real-time data of each environmental parameter with its corresponding preset safety threshold to determine the current safety margin of each environmental parameter; calculating the rate of change of each environmental parameter within a preset time window, and adjusting the corresponding current safety margin based on the rate of change to obtain a safety adjustment margin; determining the current operating condition based on the current environmental parameters, and dynamically allocating the weights of each environmental parameter under the current operating condition based on historical abnormal event data; and generating the environmental safety score based on the weights of each environmental parameter and the corresponding safety adjustment margin.

[0039] Optionally, the security linkage method also includes:

[0040] The charging compartment controller monitors multi-source operating data in real time. When any data anomaly is detected during monitoring, it serves as a trigger condition to initiate a safety process of reporting the anomaly, issuing a command from the controller, and executing a power-off interlock on the charging equipment. This safety process is accompanied by continuous data acquisition throughout. The multi-source operating data includes environmental parameters and operating parameters of all equipment inside the compartment.

[0041] In a second aspect, embodiments of this disclosure provide a charging device for providing power to a battery, the battery having wireless communication capabilities, and the charging device comprising:

[0042] A charging device, electrically connected to the battery, is used to provide electrical energy to the battery;

[0043] The charging compartment controller is capable of communicating with both the battery and the charging device, and controls the charging process of the battery through the charging device by executing the safety linkage method between the charging compartment controller and the battery as described in any of the foregoing embodiments.

[0044] Compared with the prior art, the technical solution of the present disclosure has the following advantages:

[0045] In the safety linkage method between the charging compartment controller and the battery provided in this embodiment, the charging compartment controller determines the connection quality level to ensure the reliability of the charging circuit at the physical level. Once the connection quality meets the standard, a pre-charging self-test phase is initiated, diagnosing the battery status through interaction between the battery and the controller. After confirming the battery status is normal, a pre-charging phase begins, where the battery reports its current status information to both the battery and the charging compartment controller for dual independent verification. This ensures that the charging phase is only authorized when both parties confirm that the results meet the requirements, thereby ensuring the stability and safety of the charging process. In other words, intelligent monitoring and automatic adjustment improve the safety of the charging process. Attached Figure Description

[0046] To more clearly illustrate the technical solutions in the embodiments of this disclosure 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 only embodiments of this disclosure. For those skilled in the art, other drawings can be obtained based on the provided drawings without creative effort.

[0047] Figure 1 shows a flowchart of a safe linkage method between a charging compartment controller and a battery according to an embodiment of the present disclosure;

[0048] Figure 2 shows a flowchart of generating a connection quality level according to an embodiment of the present disclosure;

[0049] Figure 3 shows a flowchart of a battery state according to an embodiment of the present disclosure;

[0050] Figure 4 shows a flowchart of generating an environmental safety score according to an embodiment of this disclosure;

[0051] Figure 5 shows a schematic diagram of the structure of a charging device according to an embodiment of the present disclosure;

[0052] Figure 6 shows a partial structural schematic diagram of a charging compartment according to an embodiment of the present disclosure;

[0053] Figure 7 shows a partial structural schematic diagram of a charging device according to an embodiment of the present disclosure. Detailed Implementation

[0054] As described in the background section, charging control is usually performed independently by the charging device or the battery management system, but insufficient coordination between the parties may lead to information delays or misjudgments.

[0055] To address the aforementioned technical issues, this disclosure embodiment comprehensively considers multiple factors such as device status, connection quality, and bidirectional communication to construct an intelligent safety linkage system, effectively improving the safety of the charging process.

[0056] Specifically, the connection quality level is determined by the charging compartment controller, ensuring the reliability of the charging circuit at the physical level. Once the connection quality meets the standard, a pre-charging self-test phase is initiated, diagnosing the battery status through interaction between the battery and the controller. After confirming that the battery status is normal, the pre-charging phase begins, where the battery reports its current status information to both the charging device and the charging compartment controller for dual independent verification. This ensures that the charging phase is only authorized when both parties confirm that the results meet the requirements, thereby ensuring the stability and safety of the charging process. In other words, intelligent monitoring and automatic adjustment improve the safety of the charging process.

[0057] In other words, this embodiment of the disclosure achieves full-chain safety coverage from physical connection to system state, and from unilateral judgment to multi-party collaboration, through closed-loop control of "connection assessment, state diagnosis, and safe execution." This method elevates traditional charging control, which relies on static threshold judgment, to a dynamic decision-making process based on multi-source information fusion, forming an intelligent safety protection mechanism that can adapt to battery state and connection conditions, significantly improving the reliability and safety of the charging process.

[0058] To enable those skilled in the art to better understand and implement this solution, the following detailed description of the specific solutions, principles, advantages, and effects of the embodiments of this disclosure is provided with reference to the accompanying drawings and specific examples.

[0059] Referring to Figure 1, which is a flowchart of a safe linkage method between a charging compartment controller and a battery in an embodiment of this disclosure, the method pre-establishes a state machine in the charging compartment controller, including plug-in status monitoring, battery status monitoring, pre-charging, charging, and abnormal handling, and dynamically adjusts and drives state switching based on the monitored battery status.

[0060] In some embodiments, the charging compartment controller controls the battery charging process via the charging device, and the battery has wireless communication capabilities. Thus, the battery can communicate with both the charging compartment controller and the charging device via its own wireless communication function.

[0061] Furthermore, the reason for equipping the battery with wireless communication capabilities is that the charging case controller acts as the "brain" of the entire safety linkage mechanism. With wireless communication, the charging case controller can proactively and continuously query the battery, send commands (such as applying test stimuli), and receive data, without relying entirely on the charging device as an intermediary. This allows the charging case controller to obtain an independent information source, perform cross-verification, and thus lead the entire safety process, avoiding the risks associated with delegating all decision-making logic to the charging device or the battery itself.

[0062] In short, adding wireless communication functionality to the battery essentially upgrades the traditional master-slave bidirectional control architecture of "charging device-battery" to a triangular collaborative control architecture of "charging compartment controller, charging device and battery".

[0063] This configuration enables key safety steps, such as independent assessment of connection quality, in-depth interactive diagnostics of battery status, and final dual independent verification, as described in this embodiment, thereby constructing a layered intelligent safety protection system. Therefore, wireless communication functionality is not an optional add-on, but rather a crucial factor in achieving this solution.

[0064] In one embodiment, wireless communication functionality can be implemented via WiFi, 4G, 5G, and potentially 6G in the future.

[0065] As shown in Figure 1, the following steps can be performed:

[0066] S101, the charging compartment controller generates a connection quality level by acquiring the connection status between the battery and the charging device.

[0067] In some embodiments, the charging device and the battery are connected via a connector, so the connection status can be determined by monitoring the state of the connector, thereby determining the connection quality level between the charging device and the battery. This reflects whether the connection between the charging device and the battery meets operational requirements, such as communication or charging requirements.

[0068] Specifically, the charging device can communicate with the battery, or the charging device can supply power to the battery through a connector.

[0069] In some embodiments, referring to the flowchart of generating a connection quality level in an embodiment of this disclosure shown in FIG2, the following steps can be performed as shown in FIG2:

[0070] S201, during the process of connecting the charging device and the battery through the connector, the charging compartment controller synchronously collects the spatiotemporal sequence data of the connection point to construct a multidimensional spatiotemporal observation dataset.

[0071] In some embodiments, when a charging device is connected to a battery via a connector, the electrical parameters of the connection point will change, thereby enabling the synchronous acquisition of spatiotemporal sequence data. This spatiotemporal sequence data includes, but is not limited to, changes in current, voltage, temperature, and at least two of the vibration data.

[0072] In one example, the voltage can be the voltage division value on the CC2 signal line.

[0073] The CC2 signal line is a pin inside the charging connector used for "connection confirmation". It is typically detected using a resistor divider circuit.

[0074] Specifically, on the charging case controller side, a fixed reference voltage is provided through a pull-up resistor, while on the battery side, there are one or more precisely defined grounding resistors. When the connector is not connected, the voltage detected by the charging case controller on the CC2 pin is either high (such as the pull-up power supply voltage) or floating. When the connector is fully engaged, the pull-up resistor on the charging case controller and the grounding resistor inside the battery form a complete voltage divider circuit. The charging case controller can read the voltage division value on the CC2 pin.

[0075] Other data can be acquired in real time through a multimodal sensor array, and aligned and fused in the spatiotemporal domain to construct a multidimensional dataset that reflects the dynamic changes in the insertion process.

[0076] By acquiring multiphysics data in real time during the insertion process, a complete dataset in both time and space dimensions is obtained, which provides a foundation for subsequent inversion solutions and causal analysis.

[0077] Next, the data from different physical sensors are synchronized, aligned, and fused to form a complete spatiotemporal observation dataset, which accurately reflects the multi-physics coupling effect during the insertion process.

[0078] S202, the charging compartment controller establishes a parameterized forward model describing the multiphysics coupling relationship of the plug-in point, and constructs a loss function based on the difference between the predicted output of the forward model and the multidimensional spatiotemporal observation dataset. Combining the attribute information of the connector, the controller uses a gradient optimization algorithm to inversely solve for the intrinsic parameter set of the plug-in point.

[0079] In some embodiments, the forward model describes the behavior of physical fields during the connection process, such as how current, pressure, and temperature change over time. This model typically involves the coupling relationships of multiple physical fields (such as electromagnetic, thermal, and mechanical fields). To make this model more accurately reflect the actual process, it needs to be parameterized, meaning that the physical parameters involved in the forward model are adjustable.

[0080] A loss function is a tool used to measure the difference between a model's predictions and actual observed data. By minimizing this loss function, the parameters of the forward model can be optimized to better reflect reality. In this process, the model's parameters are continuously adjusted and the prediction results optimized by comparing the differences between the forward model's predicted output and the multidimensional spatiotemporal data obtained from sensors.

[0081] Furthermore, to optimize this loss function, gradient optimization algorithms (such as gradient descent) can be used. This allows the parameters in the forward model to be adjusted based on the gradient information of the loss function, making the prediction results of the forward model closer to the actual observation data, and ultimately retrieving the intrinsic parameter set of the interpolation point.

[0082] In some embodiments, the forward model simulates the physical behavior during the insertion process by modeling the coupling relationship of multiple physics fields of electro-thermal-mechanical-acoustic fields, and is solved by finite element analysis (FEA) or computational fluid dynamics (CFD) methods.

[0083] In some embodiments, the intrinsic parameter set may refer to contact resistance, actual contact area, and locking force.

[0084] S203, the charging case controller establishes a causal directed acyclic graph with the physical parameters of the connector as nodes and the physical mechanism as edges based on the intrinsic parameter set, and performs virtual intervention analysis on the causal directed acyclic graph to calculate the causal effect strength of each intrinsic parameter on the connection quality and identify key causal parameters.

[0085] In some embodiments, based on the obtained intrinsic parameter set, a causal directed acyclic graph is constructed and intervention analysis is performed to calculate the impact of each intrinsic parameter on the connection quality.

[0086] A causal directed acyclic graph (DAG) is a graphical structure where nodes represent physical parameters or variables in a system, and edges represent causal relationships between these variables. The purpose of constructing a DAG is to clarify the influence relationships between different physical parameters, especially how they affect the quality of connectors.

[0087] Virtual intervention analysis simulates the impact of changing certain parameters in a causal graph on the battery. This allows for the identification of which parameters play a critical role in connection quality.

[0088] For example, if temperature changes have a significant impact on connection quality, then the temperature parameter might be a critical node in a cause-and-effect graph.

[0089] Once the causal graph is established, the strength of the causal effect of each physical parameter on the connection quality can be calculated. That is, by analyzing how each parameter affects the final connection quality through different paths, the most important parameters (key causal parameters) can be identified.

[0090] In some embodiments, the physical parameters of the connector include: loop impedance, mating displacement, contact pressure, and one or more of current and voltage.

[0091] Correspondingly, the physical mechanism refers to the laws or mechanisms governing the interaction between physical parameters during the mating process. Specifically, the physical mechanism describes the causal relationship between different physical quantities, explaining why one physical parameter affects another.

[0092] Among them, physical mechanisms can refer to current-voltage relationships, electromagnetic induction, etc.

[0093] S204, the charging case controller determines the connection quality level based on the values ​​and uncertainties of the key causal parameters.

[0094] In some embodiments, by performing the foregoing steps, it is possible to identify the key causal parameters that have the most significant impact on connection quality. These parameters are typically closely related to the physical characteristics of the connector and its behavior during the mating process.

[0095] The values ​​and uncertainties of these key causal parameters (e.g., temperature and pressure fluctuation ranges) can be used to assess connection quality. By comprehensively considering the values ​​of these parameters and their uncertainties, the quality level of the connection can ultimately be determined. The connection quality level is typically a quantitative indicator, such as excellent, good, average, or poor, reflecting the stability and reliability of the connection process.

[0096] In one example, a Bayesian inference framework is used to propagate the values ​​and uncertainties of the key causal parameters as random variables, and the confidence interval of the connection quality level is calculated through the updated posterior distribution.

[0097] Thus, through multi-dimensional spatiotemporal data acquisition, forward model optimization, and causal graph analysis, the interrelationships of various physical parameters during the connection process are deeply understood, and key factors affecting connection quality are identified through these analyses. Finally, by integrating these factors, the connector's quality level is evaluated. The goal of this process is to improve connector performance and ensure stable and safe battery connection during charging.

[0098] S102, in response to the connection quality level being greater than a preset quality level threshold, a pre-charging self-test phase is executed, and the state of the battery is determined by the charging case controller through the interaction between the charging case controller and the battery.

[0099] In some embodiments, once it is determined that the connection between the battery and the charging device meets the requirements, a pre-charging self-test phase can be further performed. During this process, the battery's condition can be checked and determined, specifically whether the battery's performance and health status meet the usage requirements.

[0100] In some embodiments, referring to the flowchart of a battery state in an embodiment of this disclosure shown in FIG3, the following steps can be performed as shown in FIG3:

[0101] S301, a wide-spectrum pulse voltage excitation is applied to the battery through the charging compartment controller, and the charging compartment controller synchronously collects the time-domain response data of the entire process.

[0102] In some embodiments, the charging case controller can control the charging device to apply a wide-spectrum pulse excitation signal to the battery, or directly to the charging device, the signal's spectrum covering multiple frequency ranges of the charging system. This pulse signal can elicit different frequency responses, providing full-band dynamic behavior data.

[0103] By synchronously acquiring voltage and current signals from the output terminal of the charging device and the input terminal of the battery, a full-domain response dataset of the charging system is obtained, including output voltage, output current, input voltage, and input current. This dataset contains the dynamic response under different excitation signals, providing basic data for subsequent analysis.

[0104] S302, based on the full-process time-domain response data, the charging compartment controller constructs a joint feature vector describing the static and dynamic characteristics of the battery.

[0105] In some embodiments, the original global response dataset is a complex mixed signal (containing the response directly caused by the excitation and the system's own oscillations), so it is necessary to perform decoupling to obtain the signals corresponding to different types of parameters, and then determine the joint feature vector describing the static and dynamic characteristics of the battery through fusion.

[0106] In one example, step S302 may include:

[0107] S3021, The time-domain response data is decomposed by the charging compartment controller to decouple the signal into a first component characterizing the forced response and a second component characterizing the inherent oscillation characteristics.

[0108] In some embodiments, Empirical Mode Decomposition (EMD) is used to perform signal decoupling. EMD is a commonly used adaptive signal decomposition method that decomposes the original global response dataset into a set of components called Intrinsic Mode Functions (IMFs), which are the frequency and time local features of the signal.

[0109] More specifically, the original global response dataset is progressively decomposed so that each component corresponds to a signal in a different frequency band, until only one residual remains. The result of each decomposition is an IMF that matches the local oscillation characteristics.

[0110] The iteration stops and the decomposition is complete when the change in the IMF component between any two iterations is less than a set threshold.

[0111] After decomposition, the frequency characteristics of each IMF component are analyzed. Generally, IMFs with higher frequencies, higher energy, and faster decay correspond to the "forced response" (i.e., the first component) directly caused by the pulse excitation. IMFs with lower frequencies and longer decaying oscillation durations correspond to the "inherent oscillation" (i.e., the second component) that reflects the mechanical and electrical connection characteristics of the system.

[0112] This adaptive decomposition allows for the separation of different physical components from the original signal.

[0113] S3022, Perform frequency domain analysis on the first component to determine the impedance spectrum of the battery, and extract the amplitude frequency response and phase frequency response to form a frequency response feature set.

[0114] In some embodiments, frequency response characteristics refer to the frequency response characteristics exhibited by the battery after an external excitation signal is applied, including the battery's impedance spectrum, phase angle, amplitude spectrum, etc.

[0115] In this process, frequency analysis techniques (such as Fourier transform) are used to perform frequency domain analysis on the first component, thereby obtaining the battery's impedance spectrum. For example, by measuring the battery's voltage and current responses, the battery's impedance can be calculated, and the amplitude-frequency response and phase-frequency response can be extracted.

[0116] Next, by converting the forced response signal to the frequency domain, amplitude and phase information are extracted. The amplitude spectrum reflects the battery's conductivity characteristics, while the phase angle reflects the energy transfer characteristics during the battery's charging and discharging process.

[0117] S3023, Based on the second component, perform modal parameter identification to extract the modal parameters of the battery and form a modal parameter set.

[0118] In some embodiments, modal parameters reflect changes in the battery's internal resistance, battery health status (such as state of health), and inherent oscillation characteristics.

[0119] In this process, the frequency response function is used to perform frequency domain analysis on the second component, thereby obtaining the modal parameters.

[0120] S3024, determine the set of interaction terms between the frequency response feature set and the modal parameter set, and filter out the interaction terms whose significance is higher than the threshold from the set of interaction terms.

[0121] In some embodiments, all interaction terms can be computed based on a set of frequency response features and a set of modal parameters. These interaction terms represent the interaction between the frequency response features and the modal parameters, which can be implemented through multiplication operations.

[0122] Furthermore, a feature selection algorithm based on mutual information is used to determine the saliency of each interaction term. For example, for each interaction term, the mutual information between the interaction term and the target variable is calculated, and interaction terms that satisfy the saliency being greater than a threshold are retained.

[0123] In one embodiment, mutual information is used to measure the correlation between interaction items and battery health. Mutual information measures the dependency between two variables; a higher value indicates a stronger correlation between the two variables.

[0124] S3025, using the interaction term, the frequency response feature set and the modal parameter set are concatenated according to the structure to form the joint feature vector.

[0125] By adopting the scheme in step S302, the frequency response characteristics, modal parameters and their interaction terms are effectively integrated to construct a comprehensive excitation response feature vector. Through saliency screening and feature interaction, rich feature information is provided for battery health monitoring, fault diagnosis and life prediction.

[0126] S303, the joint feature vector is fused with the long-term time-series operation feature vector of the battery under natural operating conditions to construct a multimodal feature tensor.

[0127] The long-term time-series operating characteristic vector of a battery under natural operating conditions is composed of multiple parameters, including: voltage, current, temperature, state of charge (SOC), state of health (SOH), internal resistance, number of charge / discharge cycles, and charging / discharging power.

[0128] In one embodiment, step S303 includes: aligning the joint feature vector with the long-term time-series running feature vector, projecting the aligned feature vector onto a unified feature space through an encoder network, and then obtaining the multimodal feature tensor through tensor concatenation.

[0129] By using feature fusion, a qualitative leap has been achieved in the ability to perceive battery health status. It elevates diagnosis from a passive "post-event judgment" based on a single symptom to a proactive monitoring solution based on multi-dimensional evidence, constructing a holographic digital profile of the battery and enhancing its robustness.

[0130] S304, the multimodal feature tensor is input into a spatiotemporal graph neural network model based on an attention mechanism. The spatiotemporal graph neural network model uses battery cells as nodes and physical connections as edges. It captures spatial dependencies through graph convolution and analyzes feature evolution trends based on temporal convolution to generate temporal evolution results.

[0131] In some embodiments, by performing steps S301 to S303, multimodal feature tensors can be determined. Based on this, spatial dependencies can be determined using a pre-built spatiotemporal graph neural network model.

[0132] At the same time, temporal convolution can generate temporal evolution results.

[0133] Specifically, the multimodal feature tensor includes various measurement data from multiple individual battery cells at different time points. This input tensor contains the following information: time dimension information, spatial dimension information, and feature dimension.

[0134] Among them, the time dimension information corresponds to the feature values ​​of all battery cells collected at each time step, the spatial dimension information corresponds to each battery cell corresponding to a node, the physical connections between batteries constitute the edges of the graph, and the feature dimension corresponds to the different features of each battery cell.

[0135] To represent the spatial dependencies between individual battery cells, these individual battery cells need to be constructed into a graph structure:

[0136] Node: Each battery cell is considered a node.

[0137] Edge: The physical connections between individual battery cells form the edges of the graph. These may include series connections, parallel connections, or other physical relationships between battery cells.

[0138] Graph Convolutional Networks (GCNs) are used to capture the spatial dependencies between individual battery cells. Specifically, the features of each node are updated by taking into account the features of its neighboring nodes.

[0139] To further improve the performance of graph convolution, the model employs a graph attention network (GAT), which introduces an attention mechanism to learn the importance of different neighbors to the current node.

[0140] The role of the attention mechanism is that different neighboring nodes have different degrees of influence on the current node, and the attention mechanism automatically adjusts the influence of neighbors by learning weights.

[0141] For example, when the voltage of a certain battery cell changes very drastically, it has a greater impact on adjacent battery cells, so it should be given a larger weight when calculating the weighted average.

[0142] In this way, the model can flexibly capture the heterogeneous relationships between individual battery cells, thereby improving prediction accuracy.

[0143] Temporal convolutional networks (TCNs) are used to model the temporal dependencies of individual battery cell features over time. For example, the features of each battery cell change continuously over time, and temporal convolution is used to extract local temporal patterns in the time series.

[0144] Temporal convolutional networks slide along the time dimension through convolution operations to capture temporal trends (such as the rising trend or periodic fluctuations of battery temperature).

[0145] Thus, by combining graph convolution and temporal convolution, spatiotemporal joint modeling can be performed, capturing the spatial dependencies between individual battery cells and the temporal evolution of features.

[0146] In other words, the various measurement data of all battery cells at different times are organized into a large tensor, and then a graph structure is constructed with battery cells as nodes and their physical connections as edges. Then, graph convolution with attention mechanism is used to model the spatial coupling relationship between battery cells, and temporal convolution is used to characterize the changing trend of these features over time. Finally, a prediction or representation of the evolution of battery cell state over time is obtained.

[0147] S305, the charging compartment controller determines the state of the battery based on the spatial dependence and the temporal evolution results.

[0148] In one embodiment, the state of the battery is determined by rule-based reasoning based on spatial dependence and temporal evolution results, such as at least one of internal short circuit, lithium plating, or aging.

[0149] For example, by analyzing the node features of a spatiotemporal graph neural network, abnormal cells in a battery pack can be identified and located, and abnormal battery behavior can be detected.

[0150] In one optional embodiment, during the process of determining the state of the battery through the interaction between the battery and the charging case controller, the battery also establishes communication with the charging device, and the charging device requests the battery to send its own state information; in response to not receiving a reply signal from the battery within a preset time, the charging process is terminated.

[0151] For example, after the charging case controller determines that the connection is intact, the battery sends its own status information to the charging device for two-way communication confirmation.

[0152] If communication fails and no information is received within a preset time, the charging process will be terminated and an alarm will be triggered.

[0153] S103, in response to the battery being in a normal state, the pre-charging stage is entered, and the charging case controller obtains the current state information from the battery and sends the current state information to the charging device.

[0154] In some embodiments, once the battery's state is determined to be normal (e.g., the battery is in a healthy state, the battery can communicate or receive power normally, etc.), a pre-charging phase is initiated. At this time, the battery can send its current state information to both the charging device and the charging compartment controller, and the charging process is executed through the coordinated action between the charging device and the charging compartment controller.

[0155] S104, in response to the confirmation results of the current status information by the charging compartment controller and the charging device both meeting the set requirements, the charging phase is entered, and the charging compartment controller controls the execution of the charging process for the battery; wherein, when the charging compartment controller detects an abnormality in the charging process, it outputs a stop command to the charging device to stop the current charging process.

[0156] Specifically, the charging device matches and evaluates the battery's required charging voltage and current with its own output capability parameters to determine whether the charging conditions are met; in response to meeting the charging conditions, a two-way communication confirmation is performed between the charging device and the battery to verify the integrity and rationality of the received information; and the charging compartment controller independently assesses charging safety based on the current status information; in response to meeting the charging conditions, successful two-way communication confirmation, and no abnormality detected by the charging compartment controller, charging is authorized to begin; in response to any abnormality, the current charging process is stopped.

[0157] In other words, this solution ensures battery safety during charging by guaranteeing information matching between the battery, the charging case controller, and other components through real-time data transmission and verification. Simultaneously, it monitors the battery's health status in real time, immediately interrupting charging and triggering an alarm in case of any abnormality, thus maximizing the safety of the charging process and protecting battery health.

[0158] It should be noted that, for any of the processes described in the above scheme, if the conditions are not met, the charging process will be stopped immediately and an alarm will be triggered.

[0159] The matching assessment may include: comparing the battery's required voltage with its own maximum output voltage capability, for example, requiring the required voltage to not exceed 95% of the maximum output voltage; comparing the battery's required current with its own maximum output current capability, for example, requiring the required current to not exceed 90% of the maximum output current; verifying that the battery's required power is within its rated power range and retaining at least 10% safety margin, etc.

[0160] Two-way communication confirmation can refer to the charging device sending a randomly generated challenge code to the battery; the battery calculating the challenge code and returning a response code; and the charging device verifying the correctness of the response code to confirm the security and reliability of the communication link.

[0161] Assessing charging safety may include: verifying whether the reported individual cell voltage is within a safe range, for example, the highest individual cell voltage does not exceed 90% of the maximum allowable charging voltage; verifying whether the temperature data is within the normal operating temperature range; and checking whether the insulation resistance value is higher than the safety threshold.

[0162] In some embodiments, the environment underground in coal mines often presents rapidly changing risk factors, such as fluctuations in combustible gas concentration, changes in temperature and humidity, and changes in ventilation conditions. Therefore, before generating a connection quality level, the process further includes: the charging compartment controller generating an environmental safety score based on collected environmental parameters within the charging compartment underground in the coal mine, and generating the connection quality level when the environmental safety score exceeds a preset safety score threshold.

[0163] The determined environmental safety score can provide a preliminary assessment of whether the current environment is suitable for charging. In other words, the environmental safety score will serve as the basis for deciding whether charging can proceed.

[0164] Specifically, referring to Figure 4, a flowchart of generating an environmental safety score according to an embodiment of this disclosure is shown. As shown in Figure 4, the following steps can be performed:

[0165] S401 collects multi-source environmental parameters, including combustible gas concentration, cabin temperature, humidity, and ventilation status.

[0166] In some embodiments, sensors distributed within the charging compartment can be used to sequentially acquire multi-source environmental parameters.

[0167] Among them, the concentration of combustible gas is the concentration of combustible gas (such as methane) commonly found in coal mines, which directly affects the risk of fire or explosion during charging; excessively high temperature may cause the battery to overheat, or even cause a fire or damage to the battery; high humidity may cause the battery to short circuit or electrical equipment to fail; ventilation status: good ventilation can effectively reduce gas concentration and temperature, and reduce the risk of fire and explosion.

[0168] Real-time monitoring of changes in the underground environment provides a foundation for safe charging.

[0169] S402, compare the real-time data of each environmental parameter with its corresponding preset safety threshold to determine the current safety margin of each environmental parameter.

[0170] In some embodiments, each environmental parameter has a preset safety threshold. By comparing the environmental parameters acquired in real time with the safety threshold, the difference between the current environmental parameter and its safety threshold can be reflected, i.e., the current safety margin.

[0171] The smaller the current safety margin, the greater the risk associated with this parameter; conversely, the larger the current safety margin, the smaller the risk associated with this parameter.

[0172] S403, within a preset time window, calculate the rate of change of each of the environmental parameters, and based on the rate of change, perform trend adjustment on the corresponding current safety margin to obtain a safety adjustment margin.

[0173] In some embodiments, the rate of change of each environmental parameter is calculated within a preset time window (e.g., an evaluation is performed at regular intervals). If a parameter changes rapidly (e.g., a sudden increase in temperature or a rapid increase in gas concentration), this may indicate an increased safety risk.

[0174] Based on these rates of change, the system will perform trend adjustments to the original safety margin, that is, predict future changes in the safety margin. The safety adjustment margin at this point is the dynamically adjusted margin based on the rate of change and trend, reflecting the current actual safety status.

[0175] S404 determines the current operating condition based on current environmental parameters and dynamically allocates the weights of each environmental parameter under the current operating condition based on historical abnormal event data.

[0176] In some embodiments, the current operating conditions are determined based on current environmental parameters. For example, in a confined space, poor air circulation may be a major risk factor; while in an open environment, wind speed and gas concentration may be more critical.

[0177] Based on historical data on unusual events (such as past safety incidents), the weights of different environmental parameters are dynamically assigned. This means that some parameters may become more important under certain circumstances, while others may be considered less important. In other words, the weight of each environmental parameter varies depending on the situation.

[0178] S405, Based on the weights of each of the environmental parameters and the corresponding safety adjustment margins, the environmental safety score is generated.

[0179] In some embodiments, an environmental safety score is generated based on the weights and safety adjustment margins of each environmental parameter. This score integrates the safety status of all environmental parameters and reflects the overall environmental safety.

[0180] In one example, an environmental safety score is generated using a weighted fusion approach, based on the weights of each environmental parameter and the corresponding safety adjustment margins.

[0181] In one example, the safety adjustment margins of all environmental parameters are discretized into multiple safety levels to construct a multidimensional discrete state space. The weights of each environmental parameter are incorporated into the state transition probability calculation to form a weighted state transition probability matrix. Based on the current environmental state and the weighted state transition probability matrix, the state probability distribution for multiple future time steps is predicted. The probability of being in a dangerous state during the prediction period is integrated over time to generate the environmental safety score.

[0182] The aforementioned scheme for determining environmental safety scores adapts to the rapidly changing risk factors in the underground mining environment. This assessment method, based on real-time data, can monitor and respond to changes in the underground environment in real time. By dynamically adjusting the weights, discrete levels, and state transition matrices of environmental parameters, it can quickly reflect changes in environmental risks and make timely responses.

[0183] Furthermore, it can reduce the risk of missed or false alarms, thereby improving overall security.

[0184] In some embodiments, the following can also be performed: the charging compartment controller monitors multi-source operating data in real time, and when any data anomaly is detected during monitoring, it serves as a trigger condition to initiate a safety process of reporting the anomaly, issuing a command from the controller, and executing a power-off interlock for the charging equipment, and this safety process is accompanied by uninterrupted data acquisition throughout; wherein, the multi-source operating data covers environmental parameters and operating parameters of all equipment in the compartment.

[0185] For example, temperature sensors installed on the plug and socket of the connector monitor the temperature data of the connector contact points; the charging case controller receives and analyzes the temperature data; based on the analysis results, it is determined whether a temperature abnormality has occurred; when a temperature abnormality is determined, a limitation or interruption operation on the charging power is triggered.

[0186] Thus, by adopting the above solution, the safety of the charging process can be monitored and evaluated in real time to ensure that the battery operates within a safe range.

[0187] The above describes in detail the safety linkage method between the charging compartment controller and the battery through some embodiments. In order to enable those skilled in the art to better understand and implement it, the corresponding equipment is also described in detail below through some embodiments.

[0188] Referring to Figure 5, which shows a schematic diagram of a charging device in an embodiment of this disclosure, the charging device 500 is used to provide power to a battery 50A, which has a wireless communication function, for example, a communication module 50B is provided inside the battery 50A.

[0189] Accordingly, the charging device 500 may include:

[0190] The charging device 510 is electrically connected to the battery 50A and is used to provide power to the battery 50A;

[0191] The charging compartment controller 520 is capable of communicating with the battery 50A and the charging device 510 respectively, and controls the charging process of the battery 50A through the charging device 510 by executing the safety linkage method between the charging compartment controller and the battery as described in any of the foregoing embodiments.

[0192] For further details regarding the battery 50A, the charging device 510, and the charging compartment controller 520, please refer to the foregoing examples.

[0193] Referring to Figures 6 and 7, where Figure 6 is a partial structural schematic diagram of a charging compartment according to an embodiment of the present disclosure, and Figure 7 is a partial structural schematic diagram of a charging device according to an embodiment of the present disclosure, as shown in Figures 6 and 7, a car is installed in the charging compartment and is charged by a mining explosion-proof and intrinsically safe lithium-ion battery power supply 63.

[0194] Furthermore, the charging compartment has a mining isolation LED light 61, and a mining intrinsically safe camera 62 installed adjacent to the mining isolation LED light 61; a first water-based fire extinguisher 64 and a second water-based fire extinguisher 67, a mining explosion-proof and intrinsically safe control box 65, and a fire blanket 66 installed adjacent to the second water-based fire extinguisher 67, as well as an automatic fire extinguishing device 68 installed on the upper side wall of the charging compartment.

[0195] Correspondingly, the charging device includes: a mine explosion-proof electric rubber-tired vehicle water pump motor 71, a transformer 72 located below the mine explosion-proof electric rubber-tired vehicle water pump motor 71; and a pump head 73 and an explosion-proof three-phase asynchronous motor 74 located adjacent to the mine explosion-proof electric rubber-tired vehicle water pump motor 71.

[0196] The charging device also includes a first terminal 75 and a second terminal 76, as well as a first and a second weldable mating surface disposed between the first terminal 75 and the second terminal 76.

[0197] In addition, a grounding point 78 and a horn 79 for early warning were also provided.

[0198] It should be noted that this solution mainly focuses on the control logic, and therefore only shows part of the structure of the charging compartment and the charging device.

[0199] While the embodiments disclosed above are provided, the present invention is not limited thereto. Any person skilled in the art can make various modifications and alterations without departing from the spirit and scope of the invention; therefore, the scope of protection of the present invention should be determined by the scope defined in the claims.

Claims

1. A method for safe linkage between a charging compartment controller and a battery, characterized in that, The charging compartment controller controls the battery charging process via the charging device. The battery has wireless communication capabilities. The safety linkage method includes: the charging compartment controller acquires the connection status between the battery and the charging device and generates a connection quality level; in response to the connection quality level exceeding a preset quality level threshold, a pre-charging self-test phase is executed, where the charging compartment controller determines the battery's status through interaction with the battery; in response to the battery being in a normal state, a pre-charging phase is entered, where the charging compartment controller acquires current status information from the battery and sends this information to the charging device; in response to both the charging compartment controller and the charging device confirming the current status information meeting set requirements, a charging phase is entered, where the charging compartment controller controls the charging process of the battery; wherein, when the charging compartment controller detects an abnormality in the charging process, it outputs a stop command to the charging device to terminate the current charging process. The charging compartment controller generates a connection quality level by acquiring the connection status between the battery and the charging device. This includes: during the connection process between the charging device and the battery via a connector, the charging compartment controller synchronously collects spatiotemporal sequence data of the connection point to construct a multidimensional spatiotemporal observation dataset; the charging compartment controller establishes a parameterized forward model describing the multiphysics coupling relationship of the connection point, and constructs a loss function based on the difference between the predicted output of the forward model and the multidimensional spatiotemporal observation dataset. Combining this with the connector's attribute information, a gradient optimization algorithm is used to inversely solve for the intrinsic parameter set of the connection point; based on the intrinsic parameter set, the charging compartment controller establishes a causal directed acyclic graph (DAG) with the connector's physical parameters as nodes and physical mechanisms as edges, and performs virtual intervention analysis on the DAG to calculate the causal effect strength of each intrinsic parameter on the connection quality, identifying key causal parameters; the charging compartment controller determines the connection quality level based on the values ​​and uncertainties of the key causal parameters.

2. The safety linkage method according to claim 1, characterized in that, The connection quality level is determined based on one or more of the following: the spatiotemporal sequence data includes at least two of the following: current, voltage, temperature, and vibration; wherein the voltage is the voltage divider value on the CC2 signal line; the intrinsic parameter set includes: contact resistance, actual contact area, and locking force; the physical parameters of the connector include one or more of the following: loop impedance, insertion displacement, and contact pressure; the determination of the connection quality level based on the values ​​and uncertainties of the key causal parameters includes: using a Bayesian inference framework, propagating the values ​​and uncertainties of the key causal parameters as random variables, and calculating the confidence interval of the connection quality level through the updated posterior distribution.

3. The safety linkage method according to claim 1, characterized in that, When the connection quality level exceeds a preset quality level threshold, a pre-charging self-test phase is executed. Through interaction between the charging compartment controller and the battery, the charging compartment controller determines the battery's state, including: applying a wide-spectrum pulse voltage excitation to the battery and simultaneously acquiring full-process time-domain response data; constructing a joint feature vector describing the battery's static and dynamic characteristics based on the full-process time-domain response data; fusing the joint feature vector with the battery's long-term time-series operational feature vector under natural operating conditions to construct a multimodal feature tensor; inputting the multimodal feature tensor into a spatiotemporal graph neural network model based on an attention mechanism. This model uses individual battery cells as nodes and physical connections as edges, captures spatial dependencies through graph convolution, and analyzes feature evolution trends based on temporal convolution to generate temporal evolution results; and determining the battery's state based on the spatial dependencies and the temporal evolution results.

4. The safety linkage method according to claim 3, characterized in that, The process of constructing a joint feature vector describing the static and dynamic characteristics of the battery based on the full-process time-domain response data by the charging compartment controller includes: performing modal decomposition on the time-domain response data through the charging compartment controller to decouple the signal into a first component characterizing the forced response and a second component characterizing the inherent oscillation characteristics; performing frequency domain analysis on the first component to determine the impedance spectrum of the battery and extracting the amplitude-frequency response and phase-frequency response to form a frequency response feature set; performing modal parameter identification based on the second component to extract the modal parameters of the battery to form a modal parameter set; and determining the frequency response feature set and the modal parameters. The set of interaction terms between sets is selected, and interaction terms with significance higher than a threshold are selected from the set of interaction terms. Using the interaction terms, the frequency response feature set and the modal parameter set are concatenated according to the structure to form the joint feature vector. The step of fusing the joint feature vector with the long-term time-series operation feature vector of the battery in natural operation state to construct a multimodal feature tensor includes: aligning the joint feature vector with the long-term time-series operation feature vector, projecting the aligned feature vector onto a unified feature space through an encoder network, and then obtaining the multimodal feature tensor through tensor concatenation.

5. The safety linkage method according to any one of claims 1, 3, and 4, characterized in that, During the process of determining the state of the battery through the interaction between the battery and the charging compartment controller, the battery also establishes communication with the charging device, and the charging device requests the battery to send its own state information. If no response signal is received from the battery within a preset time, the charging process is terminated.

6. The safety linkage method according to claim 1, characterized in that, In response to the charging compartment controller and the charging device both confirming that the current state information meets the set requirements, the charging phase begins. The charging compartment controller then controls the charging process for the battery, including: the charging device matching and evaluating its own output capability parameters with the battery's required charging voltage and current to determine if the charging conditions are met; in response to meeting the charging conditions, performing bidirectional communication confirmation between the charging device and the battery to verify the integrity and rationality of the received information; and the charging compartment controller independently assessing charging safety based on the current state information; in response to meeting the charging conditions, successful bidirectional communication confirmation, and no abnormality detected by the charging compartment controller, authorizing the start of charging; and in response to any abnormality, stopping the current charging process.

7. The safety linkage method according to claim 1, characterized in that, Before generating the connection quality level, the process further includes: the charging compartment controller generating an environmental safety score based on collected environmental parameters from the charging compartment underground in the coal mine; and generating the connection quality level when the environmental safety score is greater than a preset safety score threshold. The generation of the environmental safety score based on the collected environmental parameters from the charging compartment underground in the coal mine includes: collecting multi-source environmental parameters including combustible gas concentration, compartment temperature, humidity, and ventilation status; comparing the real-time data of each environmental parameter with its corresponding preset safety threshold to determine the current safety margin of each environmental parameter; calculating the rate of change of each environmental parameter within a preset time window, and adjusting the corresponding current safety margin based on the rate of change to obtain a safety adjustment margin; determining the current operating condition based on the current environmental parameters, and dynamically allocating the weights of each environmental parameter under the current operating condition based on historical abnormal event data; and generating the environmental safety score based on the weights of each environmental parameter and the corresponding safety adjustment margins.

8. The safety linkage method according to claim 1, characterized in that, Also includes: The charging compartment controller monitors multi-source operating data in real time. When any data anomaly is detected during monitoring, it serves as a trigger condition to initiate a safety process of reporting the anomaly, issuing a command from the controller, and executing a power-off interlock on the charging equipment. This safety process is accompanied by continuous data acquisition throughout. The multi-source operating data includes environmental parameters and operating parameters of all equipment inside the compartment.

9. A charging device, characterized in that, The charging device is used to provide power to a battery, which has wireless communication capabilities. The charging device includes: a charging unit electrically connected to the battery for providing power to the battery; and a charging compartment controller capable of communicating with both the battery and the charging unit, and controlling the charging process of the battery through the charging unit by executing the safe linkage method between the charging compartment controller and the battery as described in any one of claims 1 to 8.

Images