Lithium battery switching state current test system

Abstract

The invention relates to a lithium battery switching state current test system, and an ant colony detection module further comprises a trigger configuration unit which is used for configuring a switching detection strategy for a detection ant, configuring a pheromone set at the position of a target energy storage unit when a follower completing predation meets a corresponding switching state trigger condition, and sending the pheromone set to the target energy storage unit. The pheromone set comprises a plurality of core pheromones and guide pheromones, the guide pheromones are used for enhancing the path-finding pheromones, when the detection ant reaches the energy storage unit where the guide pheromones are located, a corresponding switching detection strategy is executed, and whether a current switching state detection task is carried out or not is triggered by analyzing the task during predation. Detection is completed through a pheromone deployment mode, intrusive influence on the predation detection model and increase of detection port load are avoided, then reliable current waveform of the energy storage unit can be accurately captured in the detection process through an ant configuration detection mode, and the detection precision is improved.

Description

Technical Field

[0001] This invention relates to the field of testing systems, and more specifically, to a lithium battery switching current testing system. Background Technology

[0002] Lithium-ion battery health status detection is a crucial aspect of current battery application technology and an indispensable part of smart grid power management for new energy storage networks. It involves not only battery life and power quality but also safety, especially in the new energy field based on battery swapping technology, where battery detection is essential. Traditional charging vehicles also incorporate battery health status detection systems within the vehicle or charging station to provide real-time feedback on the user's battery status. In large-scale applications, such as disaster recovery battery packs in data centers, detection is also indispensable. With the decreasing cost of energy storage, energy storage technology has become a vital component of new energy applications. Currently, a critical issue lies in the detection of energy storage units. In mature power supply systems, different energy storage units have varying usage and deployment conditions, resulting in different lifespans and health statuses. While adjacent energy storage units are correlated, performing detection tasks on each individual unit is time-consuming and costly. Therefore, Chinese patent CN118133874B discloses an energy storage battery status detection system based on a predator-prey model, utilizing crawler and sparrow algorithms to construct... The predator-prey model aims to efficiently capture detection information from energy storage units, improving detection efficiency while satisfying the randomness and comprehensiveness of detection. However, this detection algorithm has limitations in its application scenarios. If the energy storage unit needs to operate in a switching state for a short period, two problems arise. First, the detection ants lack information consensus, so the brief switching time prevents them from capturing or locating the detection position, leading to deviations in the detection results. Therefore, energy storage units are generally not allowed to operate in a switching state, but this cannot be avoided if other units have detection tasks and the energy storage unit is on the same loop. Second, current detection and analysis of energy storage units operating in a switching state are crucial for characterizing their health. If the switching current cannot be analyzed, the entire system needs to compensate for this through more state analysis, and some characterization features will still be missing, leading to inaccurate analysis. Simultaneous detection of each energy storage unit poses a significant challenge to the stability of the load and power supply. Furthermore, this characteristic time is short, and the overall detection method requires high hardware costs. Without detection hardware circuits that can meet the above analytical relationships, accurate analysis of each energy storage unit cannot be achieved. Summary of the Invention

[0003] In view of this, the purpose of this invention is to provide a lithium battery switching current testing system.

[0004] To solve the above-mentioned technical problems, the technical solution of the present invention is: a lithium battery switching state current testing system, including an ant colony detection module, a predation analysis module, and a detection execution device. The detection execution device includes several mimicry detection units and several detection feedback units. The mimicry detection units are coupled to the energy storage unit to be tested and are used to form a corresponding detection loop with the energy storage unit according to the corresponding detection command. The detection feedback units are used to collect the operating parameter data of the detection loop. The ant colony detection module includes an ant colony generation unit and a trigger configuration unit. The trigger configuration unit is used to configure a switching detection strategy for the detection ants. The predation analysis module is configured with a comprehensive analysis strategy. The comprehensive analysis strategy is configured with switching state trigger conditions. When a follower that has completed predation meets the corresponding switching state trigger conditions, a pheromone set is configured at the location of the target energy storage unit. The pheromone set includes several core pheromones and guiding pheromones. The guiding pheromones are used to enhance the pathfinding pheromones. When the detection ant reaches the energy storage unit where the guiding pheromone is located, the corresponding switching detection strategy is executed. The switching detection strategy includes: Step A1: Wait for other detection ants to arrive at the guide pheromone location; Step A2: Transmit the currently collected operating parameter data to the corresponding detection ant; Step A3: Move to the energy storage unit where the core pheromone corresponding to the guiding pheromone is located, and wait for other detection ants to arrive at the designated energy storage unit; Step A4: Send a switching state request to the detection and execution device and obtain the corresponding operating parameter data; Step A5: Wait for the follower to prey on you.

[0005] With this setup, firstly, without affecting the detection methods of the traditional predator-prey model, when the basic detection completion rate in a region is high, a comprehensive analysis strategy can be used to analyze each energy storage unit to determine the location of the energy storage unit as the detection target. This allows for the detection of the current switching state. The detection process involves multiple detection ants simultaneously capturing detection data. Because the current switching state time is very short, a single capture method may miss the corresponding time window and the data capture may be incomplete. Therefore, multiple detection ants are used simultaneously to ensure capture reliability. To achieve the coordination of the detection ants, core pheromones and guiding pheromones are set to ensure that the detection ants are quickly deployed to the corresponding locations. At the same time, to leave the data of the detection ants idle before detection, the original data is left at the corresponding energy storage unit location by distributing the operating parameters, avoiding the influence between data due to insufficient storage space.

[0006] Furthermore, step A2 also includes generating a repulsion pheromone using a preset pheromone inversion algorithm and configuring the repulsion pheromone at the location. The repulsion pheromone is used to suppress pathfinding pheromones. The purpose of generating the repulsion pheromone is to prevent other detection ants from entering the location after one detection ant has entered, thus ensuring that the detection ants are not disturbed when performing their detection tasks during the detection process.

[0007] Furthermore, the pheromone set includes several pheromone groups, each containing a core pheromone and several guide pheromones. When a repulsive pheromone appears on a guide pheromone, the same repulsive pheromone is configured at the location of the guide pheromone within the same pheromone group. The purpose of updating the guide pheromone is to seal off the location of the core pheromone through pheromone filtering, preventing interference from other detection ants.

[0008] Furthermore, the comprehensive analysis strategy is configured with a task evaluation model. This model generates a task evaluation vector for each energy storage unit based on the currently acquired operating parameter data. The task evaluation vector consists of a completion component, a state anomaly component, and a pheromone decay component. The completion component reflects the task completion rate at that location, the state anomaly component reflects the degree of state anomaly at that location, and the pheromone decay component reflects the pheromone decay rate in the area where that location is located. The comprehensive analysis strategy filters the task evaluation vectors of the target area centered on the target location to determine whether the location meets the triggering conditions for the switching state. The purpose of the task evaluation model is to determine whether current switching state detection is allowed. This is based on the premise that the completion rate of the basic task in the vicinity is high, and that not performing basic detection tasks in the short term would have a minimal impact on the overall detection efficiency. The task completion status of each energy storage unit is evaluated through three dimensions: completion rate, anomaly degree, and pheromone decay. A vectorized approach is used to quickly determine the matching degree of the corresponding adjacent areas, thereby confirming whether the switching state triggering conditions are met, and thus initiating the corresponding switching detection task.

[0009] Furthermore, the comprehensive analysis strategy also includes a pheromone generation sub-strategy. This sub-strategy generates the pheromone set and includes generating ant colony distribution data for follower predation locations, retrieving pheromone constraints based on the ant colony distribution data, creating several pheromone groups based on the topological relationships between energy storage units, determining the location of the energy storage unit corresponding to the core pheromone in each pheromone group, and ensuring the concentration of the core pheromone satisfies the pheromone constraints. It also generates optimal distribution information based on each task evaluation vector in the target area and the topological relationships between energy storage units. This optimal distribution information includes the location of each guiding pheromone in the pheromone group, and the concentration of the guiding pheromone satisfies the pheromone constraints. The purpose of the comprehensive analysis strategy is to determine the amount of pheromone retrieved based on the ant colony distribution data, ensuring that pheromone deployment does not affect the overall detection task, while simultaneously ensuring that the ant colony is induced to reach the designated energy storage unit location within a short time. Over-deployment of the core pheromone can be avoided through pheromone constraints. Furthermore, determining the optimal detection distribution based on the task evaluation vector ensures that energy storage units with fewer detection items or detection units with potential anomalies are not assigned detection tasks, thereby improving detection accuracy.

[0010] Furthermore, the core pheromone includes path coordination information, which configures a detection sequence. Step A4 further includes the following: when a detection ant located at the core pheromone's position works, it moves along the configured detection sequence to the corresponding energy storage unit and completes the detection action. During detection, the detection ant needs to move for data calibration to obtain an accurate current waveform. Therefore, by configuring a corresponding detection sequence for the core pheromone, the detection ants can be controlled to coordinate during the detection process.

[0011] Furthermore, step A4 also includes a collaborative current waveform generation strategy. This strategy is configured with a deviation superposition algorithm. Each detection ant independently acquires current data and generates an independent current wavelet according to its corresponding detection sequence. When the detection ant positions overlap, the deviation superposition algorithm updates the independent current wavelets corresponding to the two detection ants to determine the current deviation characteristics between the energy storage units. The corresponding independent current wavelets are then corrected based on the current deviation characteristics until the detection sequence corresponding to each detection ant has completed detection. The purpose of this step is to avoid inaccurate current capture or deviations caused by problems with the energy storage unit itself, interference from other energy storage units in the loop, or errors in the detection equipment. Therefore, by detecting multiple different positions, on the loop, or not on the loop, the waveform is analyzed to obtain the deviation characteristics, and then corrected to obtain an accurate correction result.

[0012] Furthermore, the guiding pheromone is configured with a diffusion factor and an increment factor. The diffusion factor increases the pathfinding pheromone corresponding to the energy storage unit in the diffusion vector direction over time, based on the guiding pheromone. The diffusion vector points from the location of the core pheromone to the location of the guiding pheromone. The energy storage unit is configured with a pheromone enhancement algorithm to update the value of the pathfinding pheromone at the location of the guiding pheromone using the increment factor. Through the diffusion factor and the increment factor, even when no detection ants enter within a short period, the detection ants can be guided to the corresponding location by diffusing and enhancing the pheromone, thus completing the detection action.

[0013] Furthermore, it also includes a saturation dynamic feedback algorithm. This algorithm generates a corresponding enhancement factor based on the task saturation of the detection ant, and then generates the value of the pathfinding pheromone at the location of the core pheromone based on the enhancement factor. The purpose of the saturation dynamic feedback algorithm is to filter detection ants based on their task saturation, prioritizing detection ants with higher task saturation for task matching to quickly complete the predation action.

[0014] The main technical effects of this invention are reflected in the following aspects: First, by setting it up in this way, the current switching state detection task is triggered by analyzing the task during predation, and the detection is completed by deploying pheromones, avoiding any interventional impact on the predation detection model and increasing the load on the detection port. Then, by configuring the detection ants, the detection process can accurately capture the reliable current waveform of the energy storage unit, thereby improving the detection accuracy. Attached Figure Description

[0015] Figure 1 Flowchart of the triggering process of this invention; Figure 2 : Schematic diagram of the switching detection strategy of this invention. Detailed Implementation

[0016] The specific embodiments of the present invention will be further described in detail below with reference to the accompanying drawings, so that the technical solution of the present invention can be more easily understood and mastered.

[0017] A lithium battery switching state current testing system includes an ant colony detection module, a predation analysis module, and a detection execution device. The detection execution device includes several mimicry detection units and several detection feedback units. The mimicry detection units are coupled to the energy storage unit to be tested and are used to form a corresponding detection loop with the energy storage unit based on the corresponding detection command. The detection feedback units are used to collect operating parameter data of the detection loop. The ant colony detection module includes an ant colony generation unit. The pheromone logic of the original ant colony model is configured as follows: The pheromone weight calculation formula is: , For the first The pathfinding pheromone of the detection ant that enters the movement path. The total number of tagged pathfinding pheromones. For the first The return pheromone of the detected ant markers that returned via this movement path. The total number of tagged return pheromones. The preset baseline pheromone, This is the distance impact value for the travel path, which is related to the average time it takes for the data to travel along that path. This is the path response value for the mobile path, which is related to the response efficiency of the communication unit corresponding to the mobile path. The preset pathfinding impact weights, The preset return trip impact weight, The preset baseline influence weight; The pathfinding pheromone is configured with a pathfinding attenuation factor, and the return pheromone is configured with a return attenuation factor, meaning that the detection ant will choose the optimal path based on the weighted relationship. The predator's movement model remains almost unchanged, but an induced retrieval flight factor is added. After the predator deploys the guiding pheromone and core pheromone, it retrieves the corresponding detection ant after a period of time to obtain the current waveform of the switching state. This will not be elaborated upon here.

[0018] The core of this invention lies in three strategies: first, determining whether to perform current acquisition in a switching state; second, a strategy for guiding the action of detection ants with pheromone injection; and third, a current acquisition strategy after the location is determined. The ant colony detection module also includes a trigger configuration unit, which configures a switching detection strategy for the detection ants. The predation analysis module is configured with a comprehensive analysis strategy, which includes switching state trigger conditions. When a follower that has completed predation meets the corresponding switching state trigger conditions, a pheromone set is configured at the location of the target energy storage unit. The pheromone set includes several core pheromones and guiding pheromones. The guiding pheromones enhance the pathfinding pheromones. When the detection ant reaches the energy storage unit where the guiding pheromone is located, the corresponding switching detection strategy is executed. First, the first strategy is described in detail: The comprehensive analysis strategy is configured with a task evaluation model. The task evaluation model is used to generate a task evaluation vector for each energy storage unit based on the currently acquired operating parameter data. The task evaluation vector consists of a completion component, a state anomaly component, and a pheromone decay component. The completion component reflects the task completion degree at that location, the state anomaly component reflects the degree of state anomaly at that location, and the pheromone decay component reflects the pheromone decay rate in the area where that location is located. The comprehensive analysis strategy filters the task evaluation vector of the target area centered on the target location to determine whether the location meets the switching state triggering condition.

[0019] The evaluation vector for each energy storage unit is represented as follows: The completion component is: This represents the number of completed testing tasks. The total number of tasks. This is the task weight coefficient (default 0.4).

[0020] The abnormal state components are: For the first Abnormal indicators (such as voltage fluctuation amplitude). As the indicator weight, For the total number of indicators, .

[0021] The pheromone attenuation component is: The current pheromone decay rate in the region. Standard attenuation rate, .

[0022] Switching state trigger condition judgment: When the target area meets the condition... The time (vector dot product threshold) can be used to select corresponding energy storage units as evaluation targets, and then implemented through the following method: the switching state trigger judgment is upgraded from a single vector dot product to a three-dimensional collaborative decision-making model, integrating regional similarity, outlier influence, and difference constraints. The specific logic is as follows: First, the triggering status is evaluated through three dimensions: the weight fusion formula of nested conditions: To address the trigger priority issue in complex scenarios, a nested decision model is constructed using weighted coefficients: when When this condition is met, the switching state triggering condition is considered satisfied. Preferably, the basic weight allocation relationship is as follows: the initial value of the weight can be set to the similarity weight. =0.4, weight of difference =0.3, weight of abnormal impact =0.3. , For normalization parameters (similarity normalization). ; The goal is to denormalize the degree of difference, with the lower the entropy value, the closer it is to 1. Abnormal influencing factors can be used directly. The initial threshold for conditional judgment can be set to... It can be adaptively adjusted using historical data.

[0023] Region similarity calculation: First, perform cluster analysis on the vectors: Centered on the target energy storage unit, extract the task evaluation vectors of all units within a radius R (default 5 unit spacing). Clustering is performed using the K-means algorithm, as shown in the following formula: Where: K is the number of clusters; Let be the center vector of the k-th class. The similarity threshold is determined using Euclidean distance. This is the set of task evaluation vectors obtained. If the sample proportion of the cluster to which the target unit belongs... If the conditions in the regions are similar, a simplified detection mode is triggered.

[0024] Outlier impact assessment, outlier identification formula: calculate the standard deviation matrix of vectors within the region. For each vector Calculate the Mahalanobis distance: ,in, This is the mean of the region vector. If , ( (with degrees of freedom d=3), it was determined to be an outlier.

[0025] The outlier weighting coefficient is calculated as follows: Let the number of outliers in the region be... The total number of units is Then the abnormal influence factor: in This is the abnormal weighting coefficient.

[0026] The method for calculating the difference is as follows: Calculate the information entropy of vectors within the region: ; in, For the first The percentage of samples belonging to a class. The higher the entropy value, the greater the regional variability.

[0027] Furthermore, a dynamic adjustment mechanism for the triggering conditions is implemented: this mechanism uses an adaptive threshold algorithm to dynamically adjust the threshold. Based on historical trigger data, the threshold is updated using the Exponential Moving Average (EMA), with the following update formula: ; in, The smoothing coefficient is preferably [missing information]. , To calculate the threshold in real time, The threshold is the dynamically adjusted threshold from the previous moment.

[0028] To account for the impact of mutual influence between energy storage units on the harvested current, the concept of topology relation weighting is introduced. Topology weights are added to the evaluation vectors of adjacent units, as shown in the following algorithm: , in, For the first The weight of each energy storage unit relative to the target energy storage unit. The updated weights, where For unit The adjacency coefficient with the target cell (1 for adjacent cells, 0.5 for cells separated by one, and 0 otherwise). This refers to the topological influence factor. The switching state detection is triggered by performing regional analysis on the comprehensive evaluation vector to determine whether the triggering conditions are met. Alternatively, the triggering condition can be solely determined by the magnitude of the comprehensive analysis vector. A three-dimensional evaluation system is used to reduce the false trigger rate of switching state detection. Adjustments need to be made based on the type of energy storage battery. For example, the weight of voltage platform anomalies needs to be increased for lithium iron phosphate batteries.

[0029] The second key part of this invention lies in the comprehensive analysis strategy, which focuses on the generation and configuration of pheromones. Firstly, the role of pheromones is to complete the current detection action in the switching state without affecting the detection model. The comprehensive analysis strategy also includes a pheromone generation sub-strategy, which is used to generate the pheromone set. The pheromone generation sub-strategy includes generating ant colony distribution data for follower predation locations, retrieving pheromone constraints based on the ant colony distribution data, creating several pheromone groups based on the topological relationship between energy storage units, determining the location of the energy storage unit corresponding to the core pheromone in each pheromone group, and ensuring that the concentration of the core pheromone satisfies the pheromone constraints, and generating optimal distribution information based on each task evaluation vector in the target area and the topological relationship between energy storage units. The optimal distribution information includes the location of each guiding pheromone in the pheromone group and ensure that the concentration of the guiding pheromone satisfies the pheromone constraints.

[0030] The pheromone constraint model is set as follows: To avoid pheromone overload affecting the original predation model, constraints are set as follows: ,in: For the first The concentration of each core pheromone; For the first The first core pheromone corresponding to the first The concentration of a guiding pheromone; The upper limit of pheromone concentration (default) , (Mark intensity per ant / second): The optimal distribution information generation method is as follows: A greedy algorithm is used to generate the guiding pheromone positions: an adjacency matrix is ​​constructed based on the energy storage unit topology, with weights equal to physical distance; centered on the core pheromone, select pheromone locations with a distance less than or equal to... Nodes with a default spacing of 3 units are selected as candidates; candidate nodes are sorted in descending order of task evaluation vectors, and guiding pheromones are deployed first. This improves pheromone deployment efficiency and reduces interference with the original predation model. Adjustments are needed based on the physical layout of the energy storage units; for dense arrays, it is recommended to reduce the spacing to two units.

[0031] The pheromone configuration strategy and relationships are as follows: The guiding pheromone is configured with a diffusion factor and an increment factor. The diffusion factor increases over time according to the guiding pheromone, adding the pathfinding pheromone corresponding to the energy storage unit in the diffusion vector direction. The diffusion vector points from the location of the core pheromone to the location of the guiding pheromone. The energy storage unit is configured with a pheromone enhancement algorithm to update the value of the pathfinding pheromone at the location of the guiding pheromone using the increment factor. The pheromone enhancement algorithm is configured as follows: The guiding pheromone updates the pathfinding pheromone value through the pheromone enhancement algorithm. ;in, It is an increasing factor and is positively correlated with the density of detected ants; is the diffusion factor, which decays with time and distance. It is generated only by the comprehensive analysis strategy when the switching state trigger condition is met, as shown in the following formula: in, The initial pathfinding pheromone concentration, This is the diffusion attenuation factor, which increases with time t to increase diffusion efficiency. For diffusion factor, The guide pheromone, acting as a distance value, enhances the concentration of pathfinding pheromones along the target path, forcing detection ants to converge towards the core pheromone location. This is reflected in the fact that the generation of the guide pheromone depends on the underlying weighting algorithm of the pathfinding pheromone, essentially providing targeted enhancement of the pathfinding pheromone along a specific path. Centered on the core pheromone, the guide pheromone strengthens the pathfinding pheromone along the path through a diffusion vector (from the core location to the guide location). For example, when detection ants calculate path weights, the path covered by the guide pheromone will be additionally weighted. ( To enhance the coefficient, a weight of 0.3 is preferred, increasing the probability of the path being selected. The guiding pheromone compensates for the natural decay of the pathfinding pheromone through a diffusion factor D(t): when the detection ant density is low, D(t) increases (e.g., ...). =0.6), forcibly maintaining the pathfinding pheromone concentration of the target path; after the detection ants complete their assembly, D(t) decays over time. This avoids pheromone overload interfering with the original predation model. Incremental factor algorithm: , The initial increment factor is 0.1. The ant colony density coefficient is 0.05 per ant. Let be the number of detected ants at time s.

[0032] A saturation dynamic feedback algorithm is used to generate a corresponding enhancement factor based on the task saturation of the detection ant, and then generates the pathfinding pheromone value at the location of the core pheromone based on the enhancement factor. Algorithm flow: Calculate the task saturation of the detection ant: in: To detect the task execution time of Ant i; Total running time; This is the saturation coefficient. Enhancement factor is generated: in, To detect the total number of ants, The feedback coefficient is (0.8~1.0). The core pheromone concentration is updated as follows: ;in, This represents the initial concentration. Detection ants with high task saturation experience faster response times, leading to increased overall system throughput.

[0033] The third core component is the switching detection strategy: The handover detection strategy includes: Step A1: Wait for other detection ants to arrive at the guide pheromone location; Step A2: Transmit the currently collected operating parameter data to the corresponding detection ant. Step A2 also includes generating a repulsive pheromone using a preset pheromone inversion algorithm and configuring the repulsive pheromone at the designated location. The repulsive pheromone is used to suppress pathfinding pheromones. The pheromone set includes several pheromone groups, each containing a core pheromone and several guide pheromones. When a repulsive pheromone appears on a guide pheromone, the same repulsive pheromone is configured at the location of the guide pheromone in the same pheromone group. Pheromone Inversion Algorithm: When the detection ant reaches the guide pheromone location, the formula is executed: in: To reduce the intensity of pheromones; To guide the original intensity of pheromones; The reversal coefficient is preferably set between 0.5 and 0.8 to control the suppression effect. This is the time decay coefficient, which can be set to 0.01 / s to ensure that the repulsion effect disappears quickly after the detection is completed; To detect ant dwell time, a data synchronization mechanism is implemented: a circular buffer stores operating parameters (voltage and current waveforms), and hash verification ensures data transmission integrity, with synchronization latency controlled within 50μs. Repulsion pheromones reduce the probability of subsequent detection ants entering the area, avoiding data conflicts between multiple detection ants. The effect of repulsion pheromones on pathfinding pheromones is that pathfinding pheromones are updated directly by subtracting the repulsion pheromone.

[0034] Step A3: Move to the energy storage unit where the core pheromone corresponding to the guiding pheromone is located, and wait for other detection ants to arrive at the designated energy storage unit; Step A4: Send a switching state request to the detection execution device and obtain the corresponding operating parameter data; the core pheromone includes path coordination information, which configures the detection sequence. Step A4 also includes that when the detection ant located at the core pheromone location works, it moves along the configured detection sequence to the corresponding energy storage unit and completes the detection action. Step A4 also includes a collaborative current waveform generation strategy, which is configured with a deviation superposition algorithm. Each detection ant independently obtains current data and generates an independent current wavelet according to the corresponding detection sequence. When the detection ants' positions overlap, the deviation superposition algorithm updates the independent current waves corresponding to the two detection ants to determine the current deviation characteristics between the energy storage units, and corrects the corresponding independent current waves according to the current deviation characteristics until the detection sequence corresponding to each detection ant has completed the detection. The detection ant collaborative task architecture is a multi-detection ant role division architecture; main sampling ant (Ant-M): task: perform high-frequency sampling (default 1MHz) at the target energy storage unit (such as Unit-X) and generate the basic current wavelet. Hardware calibration is performed before sampling, using the following formula: in, This is the zero bias calibration value. The gain coefficient is the baseline pheromone in the predation model. Dynamically retrieved. The current waveform before hardware calibration. The current waveform after calibration.

[0035] Taking two reference sampling ants as an example (Ant-R1, Ant-R2): Task: In adjacent energy storage units Collect reference current and generate , Synchronization mechanism: Hardware clock synchronization is used (error ≤ 10ns), and a system delay reference is generated through delay compensation using a formula: ; in, This is the communication delay compensation value, obtained by measuring round-trip time (RTT). Verification Ant (Ant-V): Task: To collect the total current at the target cell's power input. The waveform consistency was verified. Correction was achieved by constructing an error source analysis and correction strategy for the current waveform. Common errors include the following types and causes: 1. Sampling delay error, caused by ADC conversion and communication delays, which can be corrected using timestamp interpolation; 2. Equipment deviation, caused by inconsistent sensitivity of detection units, which can be corrected using deviation transfer function calibration; 3. Loop interference, caused by unbalanced coupling between adjacent energy storage units, which can be corrected using a Kalman filter algorithm; 4. Synchronization error, caused by asynchronous clocks among multiple detection ants, which can be corrected using a phase alignment algorithm. First, regarding the time difference algorithm, the timestamp interpolation correction process is delay estimation: when the detection ant sends a sampling request, it carries a timestamp. Record when receiving data Calculation delay: ,in, The request is sent using a timestamp, specifically the local time when the sampling request was sent by the detection ant. In response to the received timestamp, specifically the local time when the detection ant received the sampled data, The processing time is 100ns by default. This can be corrected by compensating for the delay time, specifically the local time when the detection ant receives the sampled data.

[0036] The second problem can be corrected using a linear interpolation formula, which is: for delayed sampling points... ,implement: in, Alignment time (default 50ns). For the calibrated current value, This is the original sampled current value.

[0037] The third problem can be corrected using a Kalman filter noise reduction model. First, define the state-space equations: State transition: The observation equation is: ;in: Let be the current and rate of change state vector, representing the current state at time k, which includes the current value and rate of change; Let be the observation vector, representing the current observation value at time k; in, , Let Q be the process noise and observation noise at time k (the covariance matrix Q is the process noise covariance, and R is the observation noise covariance).

[0038] The fourth problem can be corrected through a waveform collaborative correction execution process: Specifically, a deviation superposition correction algorithm is configured, with the loop current constraint as follows: The target unit current should meet the following requirements. Diverting current to other units. Corrected formula: in: , The influence coefficient of adjacent units (obtained through the topological relationship matrix). The main sampling ant current, , The current of the ant is used as a reference for sampling; The loop deviation compensation amount is the current deviation correction value of the loop where the target unit is located, which is determined by... The difference between the theoretical value and the actual value is calculated. This also includes an optimization method for ant task scheduling: firstly, the sampling frequency is dynamically adjusted. When the current rising edge slope is detected At that time, the sampling frequency is automatically increased to 2MHz. The sampling frequency is calculated using the following formula: ;in, Based on the sampling frequency, For frequency adjustment factor, For symbolic functions, The slope threshold is used. Next, the sampling points are sparsified using the following method: for the stationary segment... Perform sparse sampling, sampling interval formula: in, , .

[0039] The anomaly handling and resampling mechanism is as follows: The waveform reliability assessment method is as follows: Calculate the reliability index. : in, This is a standard switching state current model, specifically a standard current waveform model for lithium battery switching states. The number of sampling points. When Resampling is triggered at certain times, and the number of resampling times is determined by the time of resampling. formula: The interface design method with the original model is as follows: The pheromone interaction protocol is: corrected waveform error. As a new pheromone marker: ; in: Pheromone weight formula for waveform correction task: ; The corrected task weights can be used to adjust the movement paths of subsequent detection ants by adjusting the pheromone weights. The priority factor controls the degree to which credibility affects the task weight.

[0040] Step A5: Wait for the follower to prey. At this point, the complete switching state current waveform can be obtained. The follower position update formula is configured with a forced return command, which means that when the discoverer first releases pheromones, after a preset time, the discoverer is forced to return to that position. This ensures that the data from the detection ant can be returned as soon as possible, improving data efficiency.

[0041] Of course, the above are just typical examples of the present invention. In addition, the present invention may have many other specific embodiments. All technical solutions formed by equivalent substitution or equivalent transformation fall within the scope of protection claimed by the present invention.

Claims

1. A lithium battery switching state current testing system, comprising an ant colony detection module, a predation analysis module, and a detection execution device, wherein the detection execution device comprises a plurality of mimicry detection units and a plurality of detection feedback units, the mimicry detection units being coupled to the energy storage unit to be tested and used to form a corresponding detection loop with the energy storage unit according to the corresponding detection command, the detection feedback units being used to collect the operating parameter data of the detection loop, and the ant colony detection module comprising an ant colony generation unit, characterized in that: The ant colony detection module also includes a trigger configuration unit, which is used to configure a switching detection strategy for the detection ants. The predation analysis module is configured with a comprehensive analysis strategy, which is configured with switching state trigger conditions. When a follower that has completed predation meets the corresponding switching state trigger conditions, a pheromone set is configured at the location of the target energy storage unit. The pheromone set includes several core pheromones and a guiding pheromone. The guiding pheromone is used to enhance the pathfinding pheromone. When the detection ant reaches the energy storage unit where the guiding pheromone is located, the corresponding switching detection strategy is executed. The switching detection strategy includes: Step A1: Wait for other detection ants to arrive at the guide pheromone location; Step A2: Transmit the currently collected operating parameter data to the corresponding detection ant; Step A3: Move to the energy storage unit where the core pheromone corresponding to the guiding pheromone is located, and wait for other detection ants to arrive at the designated energy storage unit; Step A4: Send a switching state request to the detection and execution device and obtain the corresponding operating parameter data; Step A5: Wait for the follower to prey on you.

2. The lithium battery switching current testing system as described in claim 1, characterized in that: Step A2 further includes generating a repulsive pheromone using a preset pheromone inversion algorithm and configuring the repulsive pheromone at that location. The repulsive pheromone is used to suppress pathfinding pheromones.

3. The lithium battery switching current testing system as described in claim 2, characterized in that: The pheromone set includes several pheromone groups, each containing a core pheromone and several guide pheromones. When a repulsive pheromone appears on a guide pheromone, the same repulsive pheromone is configured at the location of the guide pheromone in the same pheromone group.

4. The lithium battery switching current testing system as described in claim 1, characterized in that: The comprehensive analysis strategy is configured with a task evaluation model, which is used to generate a task evaluation vector for each energy storage unit based on the currently acquired operating parameter data. The task evaluation vector consists of a completion component, a state anomaly component, and a pheromone decay component. The completion component reflects the task completion degree at that location, the state anomaly component reflects the degree of state anomaly at that location, and the pheromone decay component reflects the pheromone decay rate in the area where that location is located. The comprehensive analysis strategy filters the task evaluation vector of the target area centered on the target location to determine whether the location meets the switching state triggering condition.

5. The lithium battery switching current testing system as described in claim 4, characterized in that: The comprehensive analysis strategy also includes a pheromone generation sub-strategy, which is used to generate the pheromone set. The pheromone generation sub-strategy includes generating ant colony distribution data for follower predation locations, retrieving pheromone constraints based on the ant colony distribution data, creating several pheromone groups based on the topological relationship between energy storage units, determining the location of the energy storage unit corresponding to the core pheromone in each pheromone group, and ensuring that the concentration of the core pheromone satisfies the pheromone constraints, and generating optimal distribution information based on each task evaluation vector in the target area and the topological relationship between energy storage units. The optimal distribution information includes the location of each guiding pheromone in the pheromone group and ensure that the concentration of the guiding pheromone satisfies the pheromone constraints.

6. The lithium battery switching current testing system as described in claim 1, characterized in that: The core pheromone includes path coordination information, which configures a detection sequence. Step A4 also includes that when the detection ant located at the core pheromone location works, it moves along the configured detection sequence to the corresponding energy storage unit and completes the detection action.

7. The lithium battery switching current testing system as described in claim 6, characterized in that: Step A4 also includes a collaborative current waveform generation strategy, which is configured with a deviation superposition algorithm. Each detection ant independently acquires current data and generates an independent current wavelet according to the corresponding detection sequence. When the positions of the detection ants overlap, the deviation superposition algorithm is used to update the independent current wavelets corresponding to the two detection ants to determine the current deviation characteristics between the energy storage units, and the corresponding independent current wavelets are corrected according to the current deviation characteristics until the detection sequence corresponding to each detection ant has completed detection.

8. The lithium battery switching current testing system as described in claim 1, characterized in that: The guiding pheromone is configured with a diffusion factor and an increment factor. The diffusion factor increases the pathfinding pheromone corresponding to the energy storage unit in the diffusion vector direction according to the guiding pheromone over time. The diffusion vector points from the location of the core pheromone to the location of the guiding pheromone. The energy storage unit is configured with a pheromone enhancement algorithm to update the value of the pathfinding pheromone at the location of the guiding pheromone according to the increment factor.

9. A lithium battery switching current testing system as described in claim 1, characterized in that: It also includes a saturation dynamic feedback algorithm, which generates a corresponding enhancement factor based on the task saturation of the detection ant, and generates the value of the pathfinding pheromone at the location of the core pheromone based on the enhancement factor.

Images