Battery test scheduling method and device based on multi-objective optimization driving

Abstract

The invention relates to the technical field of battery testing, solves the problem that in the prior art, battery testing tasks cannot be scheduled in a unified mode and comprehensively balanced under multi-factor constraints, and provides a battery testing scheduling method and device based on multi-objective optimization driving. The method comprises the following steps: acquiring real-time temperature in a battery testing environment, testing real-time charging and discharging parameters of a battery and task attribute information of each task in a to-be-scheduled task sequence; constructing a multi-objective optimization model according to the decision variable, the multi-objective cost function and the objective constraint condition; inputting the real-time temperature, the real-time charging and discharging parameters and the task attribute information into a multi-target optimization model to obtain a target task sequence; and according to the target task sequence, the carrying module and the test environment module are controlled to execute test scheduling of the to-be-tested batteries on the battery test stations. According to the invention, under the constraint of multiple factors, unified scheduling and comprehensive balance of the battery test tasks are realized.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of battery testing, in particular to a battery testing scheduling method and device based on multi-objective optimization driving. BACKGROUND

[0002] The charge-discharge characteristics, cycle performance and consistency performance of lithium batteries under multiple temperature points are significantly different, so in the production, quality detection and research and development verification process, multiple temperature point charge-discharge tests in the range of-40℃ to +80℃ are usually required. A typical battery test system includes a temperature control box, multiple test stations, charge-discharge channels, and a mechanical handling device for feeding and discharging. As the test scale expands, dozens to hundreds of batteries need to be continuously switched between different stations and different temperature points, making the scheduling efficiency, temperature control energy consumption control and system power management of the test system particularly important.

[0003] The existing battery test system generally includes a temperature control box, multiple battery test stations, charge-discharge equipment, and a handling mechanism for feeding and discharging. Test tasks are usually queued according to first-in-first-out or fixed order, and the controller independently controls the temperature of the temperature control box, the movement path of the mechanical arm, and the charge-discharge equipment. Due to the separation of task scheduling logic and temperature control process, handling path planning and power load distribution, the existing technology generally has the following situations in actual operation: on the one hand, test tasks are switched between different target temperature points in disorder, causing the temperature control box to frequently raise and lower the temperature between multiple temperature points, and the temperature control process lacks overall planning, thereby causing the test cycle to be lengthened and the energy consumption level to be increased; on the other hand, the mechanical arm only optimizes the path around shortening the moving distance or reducing the single handling time, without considering the temperature state or subsequent test plan in the scheduling perspective, so that the device utilization and station turnover rhythm are difficult to form synergy with the temperature control process; in addition, the arrangement of charge-discharge tasks on the time axis is mostly static or locally controlled by rules, without planned timing distribution adjustment of tasks with different current levels and different charge-discharge states, which is easy to form instantaneous power load aggregation in some time period, causing great impact on the test power supply system or the plant distribution system, increasing the operation cost and operation risk.

[0004] Therefore, how to uniformly schedule and comprehensively balance the battery test tasks under multiple factor constraints is a technical problem to be solved. SUMMARY

[0005] Therefore, the embodiments of the present application provide a battery test scheduling method and device based on multi-objective optimization driving, to solve the problem that the prior art cannot uniformly schedule and comprehensively balance the battery test tasks under multiple factor constraints.

[0006] In a first aspect, an embodiment of the present application provides a battery test scheduling method based on multi-objective optimization driving, applied to a battery test scheduling device based on multi-objective optimization driving. The device comprises a controller, a test environment module connected with the controller, used to provide required temperature conditions for battery test, a plurality of battery test stations arranged in the test environment, and a data acquisition module used to acquire data for battery test scheduling in real time and transmit the acquired data to the controller. A carrying module is connected with the controller and used to carry batteries between the loading and unloading area and the battery test stations. The method comprises: In response to a preset scheduling trigger instruction, the data acquisition module is used to acquire real-time temperature in a battery test environment, real-time charge and discharge parameters of a test battery, and task attribute information of each task in a to-be-scheduled task sequence. A multi-objective optimization model is constructed according to a preset decision variable, a multi-objective cost function, and a target constraint condition. The real-time temperature, the real-time charge and discharge parameters, and the task attribute information are input into the multi-objective optimization model to obtain a target task sequence. According to the target task sequence, the carrying module and the test environment module are controlled to perform test scheduling on the to-be-tested batteries in each battery test station.

[0007] In an optional embodiment, the response to the preset scheduling trigger instruction and the acquisition of the real-time temperature in the battery test environment, the real-time charge and discharge parameters of the test battery, and the task attribute information of each task in the to-be-scheduled task sequence by the data acquisition module comprise: According to a preset scheduling trigger event, a scheduling trigger condition is acquired, wherein the scheduling trigger event comprises completion of a previous battery test task, addition of a new battery test task, and test time timeout. The scheduling trigger condition is monitored, and when it is monitored that the scheduling trigger condition is met, the scheduling trigger instruction is acquired. According to the scheduling trigger instruction, the test environment temperature, the charge and discharge current, and the charge and discharge voltage of the test battery are acquired by the data acquisition module to obtain the real-time temperature and the real-time charge and discharge parameters. The attribute information of each task in the to-be-scheduled task sequence is analyzed to obtain the task attribute information.

[0008] In an optional embodiment, the construction of the multi-objective optimization model according to the preset decision variable, the multi-objective cost function, and the target constraint condition comprises: The execution order of the tasks in the task sequence is taken as the decision variable. According to the preset constant parameters and initialization state parameters, total temperature control energy consumption cost functions, total test time cost functions and power penalty cost functions are constructed respectively to obtain target cost functions; According to the preset weight coefficients, the target cost functions are weighted and calculated to obtain the multi-target cost functions; According to the preset task execution frequency constraint condition, the carrying path continuity constraint condition and the power constraint condition, the target constraint condition is determined; According to the decision variable, the multi-target cost function and the target constraint condition, the multi-target optimization model is constructed.

[0009] In an optional embodiment, the total temperature control energy consumption cost functions, the total test time cost functions and the power penalty cost functions are constructed according to the preset constant parameters and the initialization state parameters to obtain the target cost functions, including: The constant parameters and the initialization state parameters are obtained, wherein the constant parameters include a unit temperature difference energy consumption constant and a power threshold, and the initialization state parameters include an initial test environment temperature, an initial carrying module position and an initial total charging and discharging power; According to the unit temperature difference energy consumption constant and the initial test environment temperature, the target temperatures of adjacent tasks in the task sequence are combined to construct the total temperature control energy consumption cost function; According to the carrying module moving time, the test environment stabilizing time and the task test time length, the total test time cost function is constructed; According to the initial total charging and discharging power, the power threshold and the total power curve in the task sequence execution process, the power penalty cost function is constructed in combination with a preset penalty coefficient; According to the total temperature control energy consumption cost function, the total test time cost function and the power penalty cost function, the target cost functions are determined.

[0010] In an optional embodiment, the total temperature control energy consumption cost functions, the total test time cost functions and the power penalty cost functions are constructed according to the preset constant parameters and the initialization state parameters to obtain the target cost functions, including: According to the task execution order in the task sequence, a first target temperature of a first task and a second target temperature of an adjacent task are obtained; According to the first target temperature, the second target temperature and the initial test environment temperature, a plurality of temperature difference values are calculated; The total temperature control energy consumption cost function is obtained by integrating each temperature difference value and the unit temperature difference energy consumption constant.

[0011] In an optional embodiment, the constructing the total test time cost function according to the moving time of the transport module, the test environment stabilization time and the task test duration includes: calculating a target moving time required by the transport module according to the initial transport module position and the battery test station position corresponding to the first task in the task sequence; calculating a total sequence moving time according to the battery test station positions corresponding to adjacent tasks in the task sequence and a preset moving time matrix; calculating a target stabilization time according to the initial test environment temperature and the target temperature corresponding to the first task; calculating a total sequence stabilization time according to the target temperatures corresponding to adjacent tasks in the task sequence; summing up the test durations of the tasks in the task sequence to obtain the task test duration; summing up the target moving time, the total sequence moving time, the target stabilization time, the total sequence stabilization time and the task test duration to obtain the total test time cost function.

[0012] In an optional embodiment, the constructing the power penalty cost function according to the initial total charging and discharging power, the power threshold and the total power curve in the task sequence execution process, in combination with a preset penalty coefficient includes: obtaining the penalty coefficient, wherein the penalty coefficient includes a first penalty coefficient and a second penalty coefficient; simulating and constructing a curve of the total power changing with time in the task sequence execution process according to the initial total charging and discharging power, the charging and discharging states of the tasks in the task sequence, the charging and discharging average power and a preset test duration to obtain the total power curve; calculating a power over-standard penalty term according to the total power curve, the first penalty coefficient and the power threshold; calculating a power variance in a preset time interval according to the total power curve; calculating a power fluctuation penalty term according to the power variance and the second penalty coefficient; summing up the power over-standard penalty term and the power fluctuation penalty term to obtain the power penalty cost function.

[0013] In an optional embodiment, the inputting the real-time temperature, the real-time charging and discharging parameter and the task attribute information into the multi-objective optimization model to obtain a target task sequence includes: calculating a real-time total charging and discharging power according to the real-time charging and discharging parameter; According to the task attribute information, real-time target temperatures, real-time test durations, real-time station positions, real-time charging and discharging states, and real-time charging and discharging average powers of tasks in the to-be-scheduled task sequence are obtained; According to the real-time charging and discharging total power, real-time target temperature, real-time test duration, real-time station position, real-time charging and discharging state, and real-time charging and discharging average power, parameter assignment updating is performed on the multi-objective optimization function in the multi-objective optimization model, to obtain an updated multi-objective optimization model; According to the updated multi-objective optimization model, a differential evolution algorithm is used to perform global search on a solution space taking the task execution order as the decision variable, and multi-objective cost function values of a plurality of candidate solutions representing different task execution orders are calculated and evaluated, to obtain a candidate task sequence. The candidate task sequence is taken as an initial solution and substituted into a preset integer programming solution model, to perform optimization solution on the task execution order under the target constraint condition, to obtain the target task sequence.

[0014] In an optional embodiment, if the current battery test scenario is a preset scheduling time delay limited scenario, the method further includes: According to the waiting time of each task in the to-be-scheduled task sequence, a temperature difference between the current test environment temperature and the real-time target temperature of each task, and a deviation between the system total power predicted according to the real-time charging and discharging total power, real-time charging and discharging state, and real-time charging and discharging average power after each task is executed and the preset ideal average power, evaluation index values respectively representing time factors, temperature factors, and power factors are determined; According to preset evaluation weight coefficients, weighted calculation is performed on the evaluation index values, to obtain priority score values corresponding to each task in the to-be-scheduled task sequence; According to the priority score values, each task in the to-be-scheduled task sequence is sorted, and target tasks are selected in order of priority score values from high to low, to generate a fast scheduling task sequence for the current scheduling period; According to the fast scheduling task sequence, the conveying module and the test environment module are controlled to test and schedule the to-be-tested batteries corresponding to each target task in the fast scheduling task sequence, to implement battery testing under the scheduling time delay limited constraint.

[0015] In a second aspect, an embodiment of the present application provides a test device, including at least one processor, at least one memory, and computer program instructions stored in the memory, when the computer program instructions are executed by the processor, the method of the first aspect in the above-mentioned embodiment is implemented.

[0016] In summary, the beneficial effects of the present application are as follows: The battery test scheduling method and device based on multi-objective optimization driving provided by the embodiment of the application, the method comprises the following steps: in response to a preset scheduling trigger instruction, acquiring real-time temperature in a battery test environment, real-time charge-discharge parameters of a test battery and task attribute information of each task in a to-be-scheduled task sequence through a data acquisition module; constructing a multi-objective optimization model according to a preset decision variable, a multi-objective cost function and a target constraint condition; inputting the real-time temperature, the real-time charge-discharge parameters and the task attribute information into the multi-objective optimization model to obtain a target task sequence; and controlling a carrying module and a test environment module to perform test scheduling on to-be-tested batteries on each battery test station according to the target task sequence. The application synchronously acquires the test environment real-time temperature, the charge-discharge state and power parameters of each battery and the target temperature, test duration, station position and other task attribute information of the to-be-scheduled task at the time of scheduling triggering, takes the task execution order as the decision variable, constructs a multi-objective optimization model which simultaneously constrains the temperature control energy consumption, test tempo and system power load, and uniformly solves and generates the task sequence on the basis of the model, so that the temperature control box temperature switching, the mechanical arm carrying movement and the charge-discharge power distribution are no longer independently decided, but are cooperatively weighed and overall arranged under the same optimization framework, and then the carrying module and the test environment module are coordinately controlled through the target task sequence in the actual execution stage, thereby relieving the problems such as temperature regulation disorder, disconnection between path planning and temperature control and charge-discharge power timing imbalance in the prior art, and realizing comprehensive scheduling and overall optimization of the battery test process. BRIEF DESCRIPTION OF DRAWINGS

[0017] In order to more clearly illustrate the technical solutions of the embodiments of the application, the drawings needed to be used in the embodiments of the application will be briefly introduced as follows, and other drawings can also be obtained by those of ordinary skill in the art without any creative effort on the basis of these drawings, and these are within the protection scope of the application.

[0018] Figure 1 is a whole process schematic diagram of the battery test scheduling method based on multi-objective optimization driving in the embodiment 1 of the application; Figure 2 is a process schematic diagram of constructing a multi-objective optimization model in the embodiment 1 of the application; Figure 3 is a process schematic diagram of respectively constructing a total temperature control energy consumption cost function, a total test time cost function and a power penalty cost function in the embodiment 1 of the application; Figure 4 is a structure schematic diagram of a test device in the embodiment 2 of the application. DETAILED DESCRIPTION

[0019] The features and exemplary embodiments of the various aspects of the present application will be described in detail below with reference to the figures and examples. It should be understood that the specific embodiments described herein are merely exemplary and are not intended to limit the present application. The present application can be implemented in ways other than those specifically described herein without departing from the spirit of the present application. The following description of the examples is merely provided to give a better understanding of the present application by showing examples of the present application.

[0020] It should be noted that the relational terms herein, such as first and second, and the like, are used solely to distinguish one from another entity or action without necessarily requiring or implying any actual relationship or order between such entities or actions. Moreover, the terms "comprises", "comprising", or any other variations thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements does not include only those elements but can include other elements not expressly listed or inherent to such process, method, article, or apparatus. Without more limitations, an element defined by an "comprising" statement is not meant to exclude the existence of additional elements of the process, method, article, or apparatus that includes the stated elements.

[0021] It should be noted that all actions of acquiring signals, information or data in the present application are carried out in compliance with the corresponding data protection regulations of the place and with the authorization given by the owner of the corresponding device.

[0022] Embodiment 1 Please refer to Figure 1 The embodiment of the present application provides a battery test scheduling method based on multi-objective optimization driving, which is applied to a battery test scheduling device based on multi-objective optimization driving. The device comprises a controller, a test environment module connected with the controller, used for providing required temperature conditions for battery test, a plurality of battery test stations arranged in the test environment, a data acquisition module used for collecting data for battery test scheduling in real time and transmitting the collected data to the controller, and a carrying module connected with the controller, used for carrying batteries between the loading and unloading area and the battery test stations. Specifically, the battery test scheduling device driven by multi-objective optimization can dynamically adjust the test sequence according to the real-time working conditions through the unified coordination of the controller to the functional modules. The test environment module includes a temperature control box located in the core area of the device, which can provide the required target temperature range for different test tasks according to the temperature setting instructions issued by the controller, and maintain a stable temperature control environment among multiple battery test stations. The test stations are used to carry the batteries to be tested and complete the charge and discharge test under the specified temperature conditions. The data acquisition module connects with the test environment module, charge and discharge channel and task management unit to collect the current temperature, voltage and current parameters of each battery and task attribute data in real time, and transmit these information to the controller to enable the scheduling logic to operate based on the latest state. The handling module includes a mechanical arm responsible for transferring the batteries between the loading and unloading area and each test station, and accurately executing the path movement according to the scheduling instructions given by the controller. The modules operate cooperatively under the centralized management of the controller, enabling the scheduling system to optimize in the dimensions of energy consumption, test cycle and power load, thereby maintaining efficient, safe and grid-friendly battery testing capability under dynamic working conditions.

[0023] The method comprises: In response to a preset scheduling trigger instruction, the real-time temperature in the battery test environment, the real-time charge and discharge parameters of the test battery and the task attribute information of each task in the task sequence to be scheduled are acquired by the data acquisition module; Specifically, when a trigger event of starting a new scheduling cycle occurs, the scheduling trigger instruction corresponding to the trigger event is acquired, for example, a certain test task ends, a new test task joins the queue or reaches a preset time point. The data acquisition module includes a sensing and communication interface connected with the temperature control box, charge and discharge equipment and task management unit, which is used to collect the current temperature of the temperature control box, the voltage and current parameters of each test channel, and the target temperature, test duration, station position, charge and discharge state and average power of the task to be scheduled. After receiving the scheduling trigger instruction, the controller drives the data acquisition module to read the temperature sensor data, real-time voltage and current of the charge and discharge channel in sequence, and calculates the current instantaneous total power by combining the preset algorithm, and reads the tasks not yet executed and their attributes from the task queue. The above information is stored in the memory in a unified format as input for subsequent optimization modeling, so as to real-time grasp the temperature control box state, battery operating condition and task demand, and provide reliable data basis for reducing temperature control energy consumption, avoiding power impact on the grid and improving test resource utilization rate.

[0024] A multi-objective optimization model is constructed according to the preset decision variables, multi-objective cost function and target constraint conditions; Specifically, the decision variables are used to depict the scheduling scheme form to be optimized, which can be in the form of task execution order, binary variables describing the relationship between tasks, etc.; the multi-objective cost function quantifies the indicators such as temperature control energy consumption, total test time and power smoothness, and introduces corresponding weight coefficients to reflect the importance of different objectives; the objective constraint conditions include that each task is executed only once, the mechanical arm movement path is continuous, the physical capacity limit of the temperature control box and the charging and discharging system, and the instantaneous power does not exceed the maximum allowable power threshold, etc. Through this modeling process, the actual battery test scheduling problem is transformed into a multi-objective optimization problem with clear objectives and boundary conditions, which is convenient for subsequent solution by intelligent algorithms. In specific implementation, the controller establishes a temperature switching energy consumption model according to the relationship between temperature change and energy consumption, constructs a total time model combining the movement time between each station of the mechanical arm, the temperature control stabilization time and the test time, and defines the super power penalty term and the power fluctuation penalty term according to the total power curve of the system, and integrates the above models into a weighted cost function, and takes the decision variables as the independent variables, and takes the task integrity constraint, the path constraint and the power constraint as the conditions, thereby forming a multi-objective optimization model. With the help of the model, the scheduling strategy is no longer dependent on empirical rules, but comprehensively considers energy consumption, efficiency and grid friendliness at the numerical level, laying a foundation for subsequent optimization.

[0025] inputting the real-time temperature, the real-time charging and discharging parameters and the task attribute information into the multi-objective optimization model to obtain a target task sequence; Specifically, the real-time temperature is used to reflect the actual running temperature of the current temperature control box, the real-time charging and discharging parameters include the voltage, current and power calculated therefrom of each test channel, and the task attribute information is used to describe the demand characteristics of the temperature resources and power resources of the tasks to be scheduled. After introducing the above data as input into the multi-objective optimization model, the model can evaluate the energy consumption, time and power cost of different task permutations and combinations under the constraints of the current system state, and then select the task sequence with better comprehensive performance. In the implementation process, the controller first arranges the temperature, total power and task attributes collected in the first step into the data structure required by the model, substitutes the temperature switching path, mechanical arm movement path and power superposition of each task under different sorting into the multi-objective cost function, generates multiple candidate sequences by using global search algorithms such as differential evolution, and iteratively filters them, then uses integer programming model to finely solve near the optimal solution, and finally obtains the task permutation with the minimum cost under the given constraints, and determines the permutation as the target task sequence. With the help of this real-time data-based solving process, the scheduling system can adaptively adjust the task order under different working conditions, reduce unnecessary temperature back-and-forth adjustment, shorten the waiting time, suppress the charging and discharging power peak value, and realize more reasonable resource allocation.

[0026] According to the target task sequence, the handling module and the test environment module are controlled to perform test scheduling on the battery to be tested at each battery test station.

[0027] Specifically, the handling module is usually a mechanical arm or other automatic handling mechanism for handling the battery between the loading and unloading area and each test station, the test environment module is mainly composed of a temperature control box for providing and maintaining a set temperature condition and a control unit thereof, and the target task sequence gives the execution order of each task and information such as target temperature, station position, etc. According to the task sequence, the controller can perform integrated scheduling on the handling and temperature control process, so that the battery loading and unloading, station switching and temperature adjustment are coordinated with each other. In specific execution, the controller parses the next task to be executed according to the target task sequence item by item, issues temperature setting instructions to the test environment module in advance, so that the temperature control box completes the temperature rise or fall and reaches stability before the target task starts, then issues path and grabbing instructions to the handling module to send the corresponding battery into the specified station or transfer between stations, and configures the working current and voltage of the charging and discharging equipment according to the charging and discharging state and average power parameters of the task, until the test of the task is completed; then the subsequent task information in the sequence is continuously called until the end of the current scheduling period. Through this execution link, the task sequence output by the optimization algorithm is strictly implemented at the physical level, the temperature switching times of the temperature control box are reduced, the mechanical arm idle stroke is compressed, and the total power curve of the system is more smooth, which supports technical effects such as energy consumption reduction, power grid impact reduction and test efficiency improvement from the system operation level.

[0028] In an optional embodiment, the response to the preset scheduling trigger instruction, the real-time temperature in the battery test environment, the real-time charging and discharging parameters of the test battery and the task attribute information of each task in the to-be-scheduled task sequence are obtained by the data acquisition module. According to the preset scheduling trigger event, the scheduling trigger condition is obtained, wherein the scheduling trigger event includes completion of the previous battery test task, addition of a new battery test task and test time timeout; Specifically, the scheduling trigger event is a service scenario condition for starting a new round of scheduling calculation, typically including the completion of the last battery test task, the addition of a new battery test task to the test queue, and the current test running time reaching or exceeding the preset timeout, which together constitute the scheduling trigger condition for defining when the current scheduling scheme needs to be updated. By pre-defining the type and determination rule of the scheduling trigger event, the controller can automatically identify whether there is a scheduling demand during the test process, thereby avoiding rigid polling at fixed time intervals. In implementation, the host computer or the scheduling controller can maintain a set of scheduling trigger event configurations, logically combine task completion signals, task queue change signals, and timer states to generate scheduling trigger conditions, and use them as the basis for subsequent monitoring and judgment. With this sub-step, the scheduling mechanism is coupled with the actual test progress and task dynamics, and the scheduling timing is more suitable for the actual working conditions, which helps to reduce invalid scheduling calculations and improve the real-time performance and resource utilization of the overall scheduling system.

[0029] Monitoring the scheduling trigger condition, and when the scheduling trigger condition is met, obtaining the scheduling trigger instruction; Specifically, by continuously monitoring the scheduling trigger condition, a scheduling trigger instruction is generated when the preset condition is met. The scheduling trigger condition can be regarded as the combination result of the events configured in the previous sub-step, for example, any in-test task ends or the number of new tasks in the queue reaches a certain threshold, and the scheduling trigger instruction is an execution signal sent to the scheduling module or controller, which is used to formally start a new data acquisition and optimization modeling process. The system can monitor the task state, queue change, and timer state in the background at a high frequency, and when the logical judgment result is true, the scheduling trigger instruction is generated and written into the internal message queue or the interrupt handling process is triggered, so that scheduling is carried out in time when the business state changes. By separating monitoring and instruction generation, continuous attention to the on-site state is ensured, and frequent triggering of large-scale calculations is avoided, which has a significant effect on improving the scheduling response speed, taking into account the calculation overhead and system stability.

[0030] According to the scheduling trigger instruction, the test environment temperature, the charge and discharge current, and the charge and discharge voltage of the test battery are collected by the data acquisition module to obtain the real-time temperature and the real-time charge and discharge parameters; Specifically, once the scheduling trigger instruction takes effect, the controller invokes the data acquisition module to collect the test environment temperature and the charge-discharge current and voltage of each test battery for one complete time to obtain the real-time temperature and real-time charge-discharge parameters for scheduling calculation. The test environment temperature is usually given by a temperature sensor arranged in the temperature control box, and the charge-discharge current and voltage are provided by a measurement module connected to the test channel. The controller can read these data through an analog quantity acquisition board card or a digital communication interface, and combine the current and voltage to perform smoothing or filtering processing, so as to improve the stability and reliability of the parameters. The instantaneous power or short-time average power of each channel can be further calculated based on the voltage and current data, which is used to describe the current system power level. By using the triggered acquisition mode, the model input data is strictly aligned with the scheduling trigger time, which can accurately reflect the real working condition of the system at the time of scheduling decision, and is helpful for the subsequent multi-objective model to more accurately evaluate the temperature control energy consumption and power distribution, and reduce the decision deviation caused by measurement lag.

[0031] The attribute information of each task in the to-be-scheduled task sequence is parsed to obtain each task attribute information.

[0032] Specifically, for the to-be-scheduled task sequence, the attribute information of each task is uniformly parsed to form a structured task attribute information set. The task attribute information usually includes target temperature, test duration, corresponding station position, charge-discharge state, and average power fields, which are used to describe the demand characteristics of a single test task for temperature resources, time resources, and power resources. The controller can read the original task description from the task management module or task queue, extract the parameter fields according to the predefined data format, perform type conversion and validity check, filter out the completed or canceled tasks, and generate a task attribute array or list for algorithm processing. After this process, all to-be-scheduled tasks are uniformly expressed at the parameter level, which is convenient for subsequent multi-objective optimization models to calculate the cost of different task combinations, and is also beneficial for batch operation when introducing temperature switching energy consumption and power penalty indicators, so as to better balance energy consumption control, power grid impact suppression, and test beat arrangement at the algorithm level.

[0033] In an optional embodiment, referring to Figure 2 , the multi-objective optimization model is constructed according to the preset decision variables, multi-objective cost function, and target constraint conditions, and includes: The execution order of the tasks in the task sequence is taken as the decision variable; Specifically, the execution order of the tasks in the task sequence is taken as a decision variable, that is, the position of each to-be-tested task in the sequence is used to depict the form of the scheduling scheme itself. The scheduling system does not directly modify the physical properties of a single task in the solving process, but indirectly affects the temperature switching path of the temperature control box, the movement path of the mechanical arm, and the superposition of the total system power at each time by adjusting the arrangement order of the tasks in the sequence. In terms of implementation, the execution order can be encoded in the form of a vector or an arrangement, for example, a sequence of length equal to the number of tasks is used to represent the execution order, and the sequence is taken as the search object of the optimization algorithm. By taking the execution order as a decision variable, the variable scale is controllable, and it is convenient for algorithms such as differential evolution and integer programming to encode, mutate and select, which is conducive to obtaining an optimal scheduling scheme that takes into account the temperature control energy consumption, test time and power smoothness under limited computing resources.

[0034] According to the preset constant parameters and the initialization state parameters, the total temperature control energy consumption cost function, the total test time cost function and the power penalty cost function are constructed respectively to obtain the target cost functions; Specifically, the total temperature control energy consumption cost function , the total test time cost function and the power penalty cost function are constructed respectively in combination with the preset constant parameters and the initialization state parameters. The constant parameters include the temperature control box unit temperature difference energy consumption constant , the system maximum allowable power threshold and the penalty coefficient for describing the power penalty intensity, and the initialization state parameters include the current temperature control box temperature, the initial position of the mechanical arm and the current total system power. The total temperature control energy consumption cost function is based on the temperature switching path corresponding to the task order, and the temperature control energy consumption is accumulated according to the relationship between the temperature difference and the energy consumption constant; the total test time cost function comprehensively considers the moving time of the mechanical arm between the stations, the time required for the temperature control box to stabilize from one target temperature to the next target temperature and the test time of each task itself; and the power penalty cost function is based on the total system power curve, and quantitatively punishes the part exceeding the power threshold and the power fluctuation degree. This multi-target modeling method makes the energy consumption, efficiency and grid friendliness have clear mathematical depiction forms, which is convenient for unified trade-off in subsequent steps.

[0035] According to the preset weight coefficient, the target cost functions are weighted and calculated to obtain the multi-target cost function; Specifically, the preset weight coefficient for dynamic priority calculation is used, wherein The above target cost functions are weighted to form a multi-target cost function which comprehensively reflects multiple targets, and the multi-target cost function needs to be minimized i.e. wherein, denotes a minimization process, and the weight coefficients are generally set according to the importance of the temperature control energy consumption, the total test time and the power smoothness in the actual application, for example, in the scene where the power grid constraint is relatively tight, the weight of the power penalty term can be appropriately increased; in the scene where the test cycle requirement is strict, the weight of the time cost can be increased. In implementation, the controller normalizes or scales the three cost functions in the same dimension or comparable scale, then multiplies them by the corresponding weight coefficients and sums them up to obtain a single scalar form of the cost value, which is used to evaluate the comprehensive advantages and disadvantages of a certain task order. Through weighted combination, the multi-objective problem is converted into a single-objective optimization problem, which is convenient for optimization algorithm to sort and select, while retaining the differentiated attention to each technical index, avoiding the scheduling result being only friendly to a single target at the expense of other key performance.

[0036] According to the preset task execution frequency constraint condition, the carrying path continuity constraint condition and the power constraint condition, the target constraint condition is determined; Specifically, the task execution frequency constraint condition, the carrying path continuity constraint condition and the power constraint condition are explicitly defined as the target constraint conditions that must be met in the optimization process. The task execution frequency constraint is used to ensure that each to-be-scheduled task appears exactly once in the final task sequence, neither missing nor repeating arrangement; the carrying path continuity constraint is used to ensure that the next carrying of the mechanical arm always starts from the last resting station, so that the generated scheduling scheme is physically executable, avoiding jump-like station switching; the power constraint condition takes the maximum allowable power threshold as the benchmark to limit or soft-constraint the total system power at any time, so as to prevent the test process from causing too large instantaneous impact on the plant power grid. These constraints are usually written in the form of inequalities or equations into the optimization model, which work together with the decision variables and the cost function to make the task sequence obtained by solving have the implementability and safety in engineering.

[0037] According to the decision variable, the multi-objective cost function and the target constraint condition, the multi-objective optimization model is constructed.

[0038] Specifically, the decision variable, multi-objective cost function and objective constraint condition are incorporated into the same mathematical framework to form a complete multi-objective optimization model. The model takes the task execution sequence as the independent variable, takes the weighted comprehensive cost function as the optimization objective, and takes the task number, carrying path and power boundary as the constraint set to form a standard form of constrained optimization problem. Based on the model, the scheduling controller can call the differential evolution algorithm for global search, and then combine the integer programming method for local fine solution, to find a task sequence with smaller comprehensive cost under the premise that all constraints are met. The model thus constructed changes the battery test scheduling from an empirical schedule to a calculable, comparable and iterative optimization process, which is conducive to maintaining an optimal energy consumption level, test efficiency and grid friendliness under different operating conditions.

[0039] In an optional embodiment, referring to Figure 3 , the total temperature control energy consumption cost function, the total test time cost function and the power penalty cost function are constructed according to preset constant parameters and initialization state parameters, to obtain each objective cost function, including: The constant parameters and the initialization state parameters are obtained, wherein the constant parameters include a unit temperature difference energy consumption constant and a power threshold, and the initialization state parameters include an initial test environment temperature, an initial carrying module position and an initial total charging and discharging power; Specifically, the constant parameters are used to depict the physical properties of the system that do not change with the task in the entire scheduling period, such as the unit temperature difference energy consumption constant which describes the energy required by the temperature control box for each 1℃ lifting, and the maximum allowed power threshold of the system which is used to limit the maximum total power allowed by the system at any time; the initialization state parameters are used to describe the instant state of the system at the beginning of scheduling, including the initial test environment temperature , the initial carrying module position and the current initial total charging and discharging power . Before starting to construct the cost function, the scheduling controller needs to read all the two types of parameters, so that the cost function can reflect the actual running basis of the current temperature control box, mechanical arm and power system. By uniformly collecting these parameters at the starting point of scheduling, the model can accurately establish the baseline temperature for energy consumption calculation, the starting position for path calculation and the initial point for power calculation, so that the subsequent cost function construction has a complete and calculable starting coordinate system.

[0040] The total temperature control energy consumption cost function is constructed according to the unit temperature difference energy consumption constant and the initial test environment temperature, in combination with the target temperature of adjacent tasks in the task sequence; Specifically, in the modeling process of temperature control energy consumption, the unit temperature difference energy consumption constant reflects the energy consumption of temperature switching, and the initial test environment temperature This refers to the initial temperature of the temperature control chamber. Correlating these two parameters with the target temperature changes between adjacent tasks in the task sequence yields the temperature switching trajectory of the temperature control chamber throughout the entire task sequence execution. The corresponding energy consumption can then be accumulated based on the temperature difference. When constructing the cost function, the controller iterates through the task sequence, calculating the difference between the target temperatures of each pair of adjacent tasks and the energy consumption constant per unit temperature difference. Multiplication and accumulation are performed, and the energy consumption due to the temperature difference from the initial temperature to the first target temperature is added to the first term of the sequence, thus forming the total temperature control energy consumption cost function. This function can quantify the amplitude and frequency of temperature changes in the temperature control chamber under different task orders, providing an intuitive characterization of energy consumption and offering comparable indicators for optimization algorithms.

[0041] The total test time cost function is constructed based on the movement time of the transport module, the stability time of the test environment, and the test duration of the task. Specifically, the total test time cost function The cost function is composed of multiple elements, including the time for the transport module to move between different workstations, the time required for the temperature control chamber to stabilize at the target temperature, and the test duration of the individual task itself. The transport time is given by the time matrix between workstations, the temperature stabilization time can be estimated based on the temperature control chamber's heating and cooling capabilities and the target temperature difference, while the test duration is a parameter inherent to the task itself. When constructing this cost function, the controller calculates the movement time between the current task workstation and the next task workstation sequentially according to the task sequence, adds the temperature stabilization time to the corresponding task's test duration, and accumulates these time components into the total test time cost function. This function reflects the time spent by core equipment and the additional waiting time caused by scheduling, and is crucial for optimizing the overall cycle time and reducing idle time.

[0042] Based on the initial total charging and discharging power, the power threshold, and the total power curve during the execution of the task sequence, and combined with a preset penalty coefficient, the power penalty cost function is constructed. Specifically, a power penalty cost function is constructed. The system needs to consider the initial total charging and discharging power, the power threshold, and the total system power curve during task execution. The power curve is derived from the superposition of the current and voltage of the current channel and is used to describe the real-time power load of the system during the execution of the scheduling sequence. The power penalty model typically considers both the portion exceeding the power threshold and the impact of power fluctuations. The penalty coefficient is used to adjust the sensitivity to power safety under different scenarios. This cost function integrates or accumulates penalties for the portion exceeding the power threshold and additionally penalizes the variance or fluctuation amplitude of the total power curve to ensure that the scheduling scheme maintains a relatively smooth and controlled power distribution during execution, thereby avoiding the problem of instantaneous peaks impacting the power grid.

[0043] Based on the total temperature control energy consumption cost function, the total test time cost function, and the power penalty cost function, determine each of the target cost functions.

[0044] Specifically, in constructing the total temperature control energy consumption cost function respectively Total test time cost function and power penalty cost function After these three cost functions, the scheduling controller encapsulates them, ensuring each cost function maintains an independent mathematical structure and can be invoked individually as the three optimization objectives of the multi-objective optimization model. This step organizes the three cost functions into standardized objective functions that can directly participate in weighted summation and solution, laying the foundation for subsequent generation of multi-objective cost functions and construction of the overall optimization model. During this process, the controller ensures that each cost function is correlated with the decision variables and can smoothly participate in model solving under constraints, enabling a complete and accurate evaluation of the performance of different task sequences in terms of energy consumption, efficiency, and power.

[0045] In an optional embodiment, constructing the total temperature control energy consumption cost function based on the unit temperature difference energy consumption constant and the initial test environment temperature, combined with the target temperatures of adjacent tasks in the task sequence, includes: Based on the task execution order in the task sequence, obtain the first target temperature of the first task and the second target temperature of the adjacent task; Specifically, the task execution order determines the target temperatures that the temperature control chamber needs to reach sequentially during testing. Therefore, when constructing the temperature control energy consumption model, it is necessary to first extract this temperature information based on the task sequence. The target temperature of the first task constitutes the first temperature control target of the temperature control chamber from its initial state, and the target temperatures of adjacent tasks constitute the subsequent nodes of the temperature switching path. After reading the task sequence, the controller extracts the target temperature of the first task as the first target temperature according to the sequence order, and then extracts the target temperatures of each pair of adjacent tasks in the sequence as the second target temperature, so that the temperature change path forms a mathematically continuously calculable chain. Through this process, the temperature change trajectory of the temperature control chamber when executing different task schemes is clearly defined, providing a structured input for subsequent energy consumption calculations, and enabling the energy consumption differences caused by different temperature jump amplitudes between different task sequences to be quantified.

[0046] Based on the first target temperature, the second target temperature, and the initial test environment temperature, multiple temperature differences are calculated. Specifically, after understanding each node in the temperature path, the controller needs to calculate the temperature difference between the initial temperature and the first target temperature, as well as the temperature difference between two adjacent target temperatures. The initial test ambient temperature describes the actual temperature state of the temperature control chamber and is the starting point for temperature control energy consumption calculation; the first and second target temperatures correspond to the temperature target points in the sequence. The controller calculates the difference between the initial temperature and the first target temperature according to the temperature path sequence, and then calculates the difference between adjacent target temperatures segment by segment, thus forming multiple temperature difference arrays. Temperature differences can be positive or negative, and energy consumption occurs regardless of whether the temperature rises or falls. By mapping the real-world task sequence to continuous temperature differences, the model can accurately reflect the temperature fluctuations of the temperature control chamber under different task sequences, enabling energy consumption assessment to be based on the actual temperature change range.

[0047] The total temperature control energy consumption cost function is obtained by integrating the temperature difference values ​​and the unit temperature difference energy consumption constant.

[0048] Specifically, temperature control energy consumption is essentially linearly related to the temperature change. Therefore, after obtaining the temperature difference values ​​for each segment, multiplying and summing them segment by segment with the constant energy consumption per unit temperature difference yields the total temperature control energy consumption cost function. The integral operation here describes the summation of energy consumption over all temperature difference segments, covering the temperature control energy consumption from the initial temperature to the first target temperature and between all adjacent tasks. If the temperature change is large or the number of temperature jumps is high, the corresponding cumulative energy consumption value is higher, thus being assigned a higher cost weight during the optimization process. The specific calculation formula is as follows: in, This represents the target temperature for the j-th task; Indicates the j-th task adjacent to the j-th task. The target temperature for each task In the calculation of temperature difference, the first Temperature switching in segments. Through this calculation process, the system can fully quantify the impact of task sequences on temperature control energy consumption, making the advantages and disadvantages of different scheduling schemes in terms of temperature control energy consumption comparable, and providing a clear energy consumption evaluation basis for optimization algorithms.

[0049] In an optional embodiment, constructing the total test time cost function based on the movement time of the transport module, the stabilization time of the test environment, and the duration of the task test includes: Based on the initial position of the transport module and the position of the battery testing station corresponding to the first task in the task sequence, the target movement time required for the transport module is calculated. Specifically, the initial position of the transport module is usually the standby station or center position set by the system, which is the actual position of the robotic arm at the start of the scheduling cycle; the battery test station position corresponding to the first task represents the first operation target position of the sequence to be executed in this scheduling. After acquiring the task sequence, the controller reads these two position parameters and calculates the movement time from the initial position to the station based on the preset movement time matrix or geometric path model. Since this movement occurs at the very beginning of the scheduling sequence, there is no preceding task available for reuse, so its time cost directly determines the starting delay of the entire test sequence. By explicitly calculating this initial movement time, the total test time model can accurately reflect the necessary consumption required for the robotic arm to enter the working state, providing a reliable reference for the actual test cycle.

[0050] The total sequence movement time is calculated based on the battery test station positions corresponding to adjacent tasks in the task sequence and the preset movement time matrix. Specifically, the change in workstation position between adjacent tasks determines the path consumption of the robotic arm in the entire scheduling sequence. Therefore, it is necessary to sum the values ​​of each adjacent task pair segment by segment based on a preset movement time matrix. The movement time matrix is ​​usually predetermined by the spatial layout between workstations and the robotic arm's mobility, and can directly provide the shortest movement time between any two workstations. The controller reads the workstation position of each pair of adjacent tasks sequentially according to the task sequence, retrieves the corresponding movement time from the matrix, and accumulates them to form the total sequence movement time. This calculation process can completely record all the displacements of the robotic arm during the execution of the task sequence, enabling the scheduling model to accurately compare the impact of different task orders on the robotic arm's motion burden, and thus reduce unnecessary migrations and improve overall execution efficiency by adjusting the task order.

[0051] The target stabilization time is calculated based on the initial test environment temperature and the target temperature corresponding to the first task. Specifically, before executing the first task, the temperature control chamber needs to adjust from the current actual temperature to the target temperature of the first task. Therefore, the difference between the initial test ambient temperature and the first target temperature determines the time required for temperature stabilization. Based on the heating or cooling capacity of the temperature control chamber, the temperature difference can be converted to obtain the corresponding stabilization time, which is then used as the necessary preparation time for the temperature control system before the start of the test sequence. Since the temperature adjustment from the initial point to the first target temperature cannot be avoided through task sequencing, this time overhead also has a fundamental impact on the overall test rhythm. By explicitly calculating this stabilization time, the scheduling model can accurately present the time cost required for the temperature control system to enter the working state of the first task, providing a clear starting point for the subsequent accumulation of stabilization times based on the temperature changes of adjacent tasks.

[0052] The total sequence stabilization time is calculated based on the target temperatures corresponding to adjacent tasks in the task sequence. Specifically, during the execution of the task sequence, the temperature control system needs to adjust the temperature from the target temperature of the previous task to the target temperature of the next task between every two adjacent tasks. Therefore, the temperature difference between the target temperatures of adjacent tasks determines the required temperature stabilization time for that segment. After reading the task sequence, the controller compares the target temperatures of each pair of adjacent tasks in the sequence, converts the temperature difference into the corresponding temperature stabilization time based on the temperature control chamber's heating and cooling capabilities, and accumulates these segments to form the total sequence stabilization time. Since the temperature stabilization time has a linear or near-linear relationship with the temperature difference, different task orders will significantly change the amplitude and number of temperature jumps, thus directly affecting the total test time. Therefore, through this calculation, the scheduling model can accurately evaluate the temperature control system delay caused by task ordering, providing a quantitative basis for subsequent optimization algorithms to reduce unnecessary round-trip adjustments in the temperature dimension.

[0053] The test duration of each task in the task sequence is summed to obtain the test duration of the task. Specifically, task test duration refers to the inherent execution time required for each battery test task, including the duration of the entire charge-discharge cycle or other pre-defined test procedures. It is an incompressible time component during scheduling. The controller iterates through the task sequence, extracts the preset test duration for each task sequentially, and sums them up to form the total task test duration. This portion of time is not affected by the scheduling order, but it must be combined with the handling time and temperature stabilization time to constitute the total test time. Therefore, summing it separately in the cost function allows the model to clearly present the magnitude of the inherent workload of each task, providing a complete basis for the composition of the overall time cost. By summing all task durations, the scheduling system can accurately reflect the unavoidable basic cycle of the test process, which is a necessary component for subsequent total time optimization.

[0054] The total test time cost function is obtained by summing the target movement time, total sequence movement time, target stabilization time, total sequence stabilization time, and task test duration.

[0055] Specifically, after calculating the aforementioned time parameters, the controller summarizes the target movement time of the initial handling phase, the total sequence movement time formed by the cumulative displacement of the robotic arm throughout the entire test, the target stabilization time required for the temperature control system to enter the first task, the total sequence stabilization time formed by the temperature switching of each adjacent task, and the test duration of all tasks themselves. These time components together constitute a complete total test time cost function, providing the scheduling model with a core indicator for measuring the differences in execution efficiency among different task sequences. The calculation method for the total test time comprehensively covers all time-consuming aspects of the task execution process, enabling the scheduling system to consider not only the execution time of the tasks themselves but also the additional delays caused by the handling and temperature control systems when evaluating different task arrangements. This helps generate a scheduling scheme with a more compact overall execution cycle and less idle time.

[0056] In an optional embodiment, constructing the power penalty cost function based on the initial total charging and discharging power, the power threshold, and the total power curve during the task sequence execution, combined with a preset penalty coefficient, includes: Obtain the penalty coefficient, wherein the penalty coefficient includes a first penalty coefficient and a second penalty coefficient; Specifically, the penalty coefficient is used to adjust the strength of the power penalty in the overall cost function, where the first penalty coefficient... The first penalty applies to the portion of power exceeding the threshold, and the second penalty coefficient... The penalty is applied based on the degree of power fluctuation, and both the penalty and the penalty coefficient reflect the system's emphasis on grid security and power smoothness. Before constructing the power penalty cost function, the controller needs to read these two penalty coefficients from the parameter configuration or pre-set reasonable values ​​according to different test scenarios for subsequent weighting of excess power and power variance. Once the penalty coefficients are determined, they provide an adjustable means for soft control of power constraints, enabling the scheduling algorithm to tolerate short-term power approach to the threshold when necessary, and to increase the penalty intensity in scenarios more sensitive to grid impacts, thereby enhancing the weight of power constraints in the overall cost and achieving adaptive matching for different application requirements.

[0057] Based on the initial total charging and discharging power, the charging and discharging state of each task in the task sequence, the average charging and discharging power, and the preset test duration, the curve of the total power changing with time during the execution of the task sequence is simulated and constructed to obtain the total power curve. Specifically, the total power curve This method describes the change of the total charging and discharging power of the system over time under a given task sequence, serving as the basis for subsequent judgments on whether the power exceeds the limit and for assessing power fluctuations. Given the initial total charging and discharging power, the controller combines the charging and discharging status, average charging and discharging power, and corresponding preset test durations of each task in the task sequence to add or subtract tasks added or ended within each time period, thereby simulating the trajectory of the total power over time throughout the entire scheduling cycle. For charging tasks, the average power is included in the total power in the positive direction; for discharging tasks, it is treated as a power contribution in the corresponding direction and added according to the actual metering method, so that the total power at any given time can be obtained by summing the power of the tasks under test. Through this task-attribute-based power curve construction method, the peak power position, duration, and overall load level under different scheduling schemes can be finely characterized without increasing the additional hardware measurement burden, providing accurate input for the subsequent calculation of power penalty terms.

[0058] The power over-limit penalty item is calculated based on the total power curve, the first penalty coefficient, and the power threshold. Specifically, the power over-limit penalty term is used to quantify the degree and duration of the total power exceeding the power threshold during the execution of the task sequence, thereby suppressing scheduling schemes that may impact the power grid during optimization. After obtaining the total power curve, the controller performs discrete or continuous scanning of the time axis throughout the entire scheduling cycle, comparing the total power value at each time point with the power threshold. The portion exceeding the threshold is defined as the over-limit power interval, and this is accumulated or integrated using the first penalty coefficient to obtain a penalty value reflecting the severity of the over-limit. The larger the over-limit magnitude and the longer the duration, the higher the penalty value, and the corresponding cost in the multi-objective cost function also increases. This guides the optimization algorithm to prioritize tasks with lower power peaks and lower over-limit risks, making the scheduling results more in line with the requirements of power grid friendliness and safe operation.

[0059] Calculate the power variance within a preset time interval based on the total power curve; Specifically, power variance measures the degree of fluctuation in the total power curve during scheduling and is an important indicator reflecting system power stability. Once the task sequence is determined, the total power curve is clear. Therefore, the controller can divide the entire scheduling cycle into continuous or discrete time intervals and calculate the variance of the power values ​​within these intervals using statistical methods. A larger power variance indicates more drastic power fluctuations and more pronounced peak-to-valley differences during task execution, potentially causing unnecessary dynamic shocks to the power grid. By calculating the variance, the model can quantify the smoothness of the power curve, enabling the scheduling algorithm to identify and avoid schemes with excessive fluctuations, thereby guiding the overall power to become smoother and contributing to improved system safety and power consumption coordination.

[0060] The power fluctuation penalty term is calculated based on the power variance and the second penalty coefficient. Specifically, after obtaining the power variance, the controller weights the variance according to a preset second penalty coefficient to form a power fluctuation penalty term. The second penalty coefficient essentially adjusts the importance of power fluctuations in the overall objective cost, and can be set according to grid load characteristics or power smoothing requirements in different scenarios. By multiplying the power variance by the second penalty coefficient, the statistics originally used to describe the fluctuation amplitude can be transformed into penalty values ​​that can directly participate in the optimization model, making the model more sensitive to the severity of power changes when evaluating different task sequences. This approach effectively suppresses task combinations under high fluctuation conditions, promoting the generation of scheduling schemes that better meet grid-friendly requirements.

[0061] The power over-limit penalty term and the power fluctuation penalty term are summed to obtain the power penalty cost function.

[0062] Specifically, after obtaining the power exceedance penalty term and the power fluctuation penalty term respectively, the controller directly sums the two to form the final power penalty cost function, thus unifying the exceedance and fluctuation problems within the same function. Since the exceedance penalty term reflects the severity of the system exceeding power constraints, and the fluctuation penalty term reflects the stability of the power curve, their summation achieves comprehensive constraints on grid-friendliness requirements. The specific calculation formula for the power penalty cost function is as follows: in, This represents the power fluctuation penalty term, also known as power variance, which measures the degree of power fluctuation. The power over-limit penalty term is used to calculate the extent of power over-limit. The resulting power penalty cost function can incorporate the system's power security risk and operational stability into the comprehensive evaluation criteria when evaluating different scheduling schemes. This provides a complete and quantifiable power-level basis for multi-objective optimization, enabling the task sequence generated by the scheduling algorithm to meet testing requirements while maximizing the balance between grid security and system stability.

[0063] In an optional embodiment, inputting the real-time temperature, real-time charge / discharge parameters, and task attribute information into the multi-objective optimization model to obtain the target task sequence includes: The total real-time charging and discharging power is calculated based on the real-time charging and discharging parameters. Specifically, real-time charging and discharging parameters typically include the current and voltage data currently collected from each test channel. After obtaining these parameters, the controller sums the products of the voltage and current of each channel to obtain the total real-time charging and discharging power at the current moment. This total power reflects the instantaneous load level of the system on the power grid under the current task execution state and is the fundamental quantity for subsequent judgments on whether the power is close to or exceeds the power threshold and for evaluating the power friendliness of the scheduling scheme. By converting real-time charging and discharging parameters into a total power scalar, the scheduling model can accurately grasp the current operating conditions in the power dimension, incorporate the actual load state into the optimization process, and ensure that subsequent task sequencing not only considers the requirements of the tasks to be executed but also takes into account the power background formed by the tasks already being executed, thereby improving the consistency between the scheduling results and the on-site operating state.

[0064] Based on the task attribute information, obtain the real-time target temperature, real-time test duration, real-time workstation location, real-time charging and discharging status, and real-time average charging and discharging power of each task in the task sequence to be scheduled. Specifically, task attribute information needs to be updated in real time based on the current moment and the latest task queue. Therefore, the controller obtains the real-time target temperature, real-time test duration, real-time workstation location, real-time charging / discharging status, and real-time average charging / discharging power of each task in the task sequence to be scheduled, according to the latest task attribute information. These real-time parameters can be understood as a description of the immediate requirements of each task for the temperature control system, test time, and power resources if it were scheduled for execution at the current scheduling moment. For example, if some tasks have already been partially executed in the previous scheduling cycle, their remaining test duration and subsequent power levels need to be recalculated. By organizing and updating the real-time attributes of each task in this way, the task data used by the optimization model can accurately reflect the current remaining workload and resource requirements, making the subsequent solution results more closely match the actual test progress, rather than making decisions based on initial static task parameters. This is beneficial for improving the overall scheduling accuracy and practicality.

[0065] Based on the real-time total charging and discharging power, real-time target temperature, real-time test duration, real-time workstation position, real-time charging and discharging status, and real-time average charging and discharging power, the parameters of the multi-objective optimization function in the multi-objective optimization model are updated to obtain the updated multi-objective optimization model. Specifically, the real-time total charging and discharging power, along with the real-time target temperature, real-time test duration, real-time workstation position, real-time charging and discharging status, and real-time average charging and discharging power of each task, constitute the key input parameters of the multi-objective optimization model. Based on these parameters, the controller updates the parameters of the multi-objective optimization function in the model, recalculating cost functions such as energy consumption, time, and power penalties based on the current system state. The total temperature control energy consumption cost function adjusts the temperature switching path according to the latest temperature status and task temperature requirements; the total test time cost function re-estimates the time cost based on the spatial relationship between the current robotic arm position and the task workstation, as well as the remaining test duration; and the power penalty cost function updates the simulation results of the total power curve using the real-time total power and the average power of each task. Through this overall parameter update process, the originally statically constructed multi-objective optimization model is transformed into a dynamic model reflecting the actual working conditions at the current moment, enabling subsequent optimization processes to be evaluated based on the latest state and enhancing the adaptability of the scheduling strategy to environmental and task progress changes.

[0066] Based on the updated multi-objective optimization model, the differential evolution algorithm is used to perform a global search on the solution space with the task execution order as the decision variable, and the multi-objective cost function values ​​of multiple candidate solutions representing different task execution orders are calculated and evaluated to obtain a candidate task sequence. Specifically, after updating the parameters of the multi-objective optimization model, the model can accurately characterize the temperature control energy consumption, testing time, and power characteristics at the scheduling moment. At this point, the differential evolution algorithm is used to perform a global search across the entire solution space, searching for all possible combinations of the decision variable "task execution order." The differential evolution algorithm iteratively updates these sorting schemes by randomly initializing multiple task ordering schemes and combining mechanisms such as mutation, crossover, and selection, ensuring that each generation of candidate solutions evolves towards lower costs. Each candidate sorting is input into the optimization model as an independent solution, and the model calculates its multi-objective cost function value to reflect the overall performance of the sorting in terms of energy consumption, time, and power. By evaluating the value of each candidate generation, the algorithm can select candidate task sequences that outperform other solutions, enabling the scheduling scheme to avoid local optima traps globally and providing a high-quality initial solution foundation for subsequent refined solutions.

[0067] The candidate task sequence is substituted into a preset integer programming solution model as the initial solution to optimize the task execution order under the target constraints, thereby obtaining the target task sequence.

[0068] Specifically, after the differential evolution algorithm obtains the candidate task sequence, the scheduling system further inputs it as the initial solution into the integer programming solution model. The integer programming model can perform more refined local optimization of the task execution order while strictly adhering to the objective constraints, and is particularly suitable for handling strong constraints such as constraints on the number of task executions, the continuity of transport paths, and power constraints. During the solution process, the model further searches the feasible solution space near the candidate sequence and fine-tunes or rearranges the task order to achieve better performance on the comprehensive cost function in the final obtained task sequence. Since the initial solution originates from the global search stage and is of high quality, integer programming can quickly find high-quality optimal or near-optimal solutions that satisfy the constraints within a smaller search range, ultimately forming a target task sequence that can be directly used for scheduling execution. This sequence takes into account minimizing temperature control energy consumption, improving execution efficiency, and power stability control, providing a reliable decision-making basis for the actual operation of the scheduling system.

[0069] In an optional embodiment, if the current battery test scenario is a preset scheduling latency-constrained scenario, the method further includes: Based on the waiting time of each task in the task sequence to be scheduled, the temperature difference between the current test environment temperature and the real-time target temperature of each task, and the deviation between the total system power predicted after each task is executed based on the real-time total charging and discharging power, real-time charging and discharging status and real-time average charging and discharging power and the preset ideal average power, the evaluation index values ​​used to characterize the time factor, temperature factor and power factor are determined respectively. Specifically, in test scenarios with limited scheduling latency, the system needs to make scheduling decisions within a short time. Therefore, evaluation indexes for time, temperature, and power factors are constructed by utilizing the waiting time of each task in the sequence of tasks to be scheduled, the temperature difference between the current test environment temperature and the real-time target temperature of each task, and the deviation between the predicted total system power (based on real-time total charging / discharging power, real-time charging / discharging status, and real-time average charging / discharging power) and the preset ideal average power. A longer waiting time indicates a greater risk of the task being delayed; a smaller temperature difference indicates a lower temperature adjustment cost for the task at the current temperature; and a prediction of the total power closer to the ideal average power indicates that the task is more conducive to maintaining a stable power curve. In this way, the priority of tasks in terms of timeliness, temperature control cost, and power friendliness can be quantified without complex global optimization, forming an evaluation basis suitable for rapid scheduling.

[0070] Based on the preset evaluation weight coefficients, the values ​​of each evaluation index are weighted and calculated to obtain the priority score value corresponding to each task in the task sequence to be scheduled. Specifically, after obtaining the evaluation indicators for time, temperature, and power factors, the three indicators are weighted and calculated using preset evaluation weight coefficients to obtain the priority score value corresponding to each task in the task sequence to be scheduled. The evaluation weight coefficients can be set according to the focus of the current test scenario. For example, the weight of the time factor can be appropriately increased when the scheduling delay constraint is more stringent, and the weight of the power factor can be appropriately increased when the grid constraint is more sensitive. The calculation formula for the priority score value is as follows; in, This represents the priority score of the i-th task in the task sequence; , These represent the evaluation weight coefficients corresponding to the time factor, temperature factor, and power factor, respectively. This represents the waiting time of task i, which is the accumulated waiting time from when the task entered the task queue to the current scheduling time.

[0071] The longer the wait, the higher the priority. This indicates the temperature difference between the current test environment temperature and the target temperature of task i. The smaller the temperature difference, the less temperature control adjustment is required to perform the task, and the lower the cost. The charge / discharge state factor of task i is generally taken as +1 (charging) or -1 (discharging), and is used to reflect the impact on the direction of total power after the task is executed. This represents the preset ideal average power, used to guide the entire scheduling process to maintain a relatively stable power load that is neither excessively high nor low. This indicates the power-friendliness of task i. Through weighted scoring, the metrics of different tasks across the three dimensions are uniformly mapped to a single priority value, enabling the scheduling system to use an intuitive and comparable score to measure the urgency and scheduling value of each task, providing a direct basis for subsequent rapid sorting and selection.

[0072] Based on the priority score values, each task in the sequence of tasks to be scheduled is sorted, and target tasks are selected in order of priority score from high to low to generate a fast scheduling task sequence for the current scheduling cycle. Specifically, after obtaining the priority scores of each task, the system sorts the tasks in the task sequence from highest to lowest score, and then selects the tasks with higher scores to construct the fast scheduling task sequence for the current scheduling period. High-priority tasks are placed at the beginning of the sequence for priority execution, while low-priority tasks are deferred to later tasks. This ensures that, under limited scheduling latency constraints, tasks with longer waiting times, better temperature matching, and more favorable power curves are processed promptly. This sorting process replaces complex global searches and integer programming with simple rule-based operations, significantly reducing the computational load while retaining a comprehensive consideration of time, temperature, and power, making it suitable as a fast scheduling strategy for latency-sensitive scenarios.

[0073] According to the fast scheduling task sequence, the transport module and the test environment module are controlled to perform test scheduling on the batteries to be tested corresponding to each target task in the fast scheduling task sequence, so as to achieve battery testing under the constraint of scheduling delay.

[0074] Specifically, after the rapid scheduling task sequence is determined, the controller sequentially controls the transport module and the test environment module according to the sequence to perform specific test scheduling operations on the batteries to be tested corresponding to each target task. This includes driving the transport module to deliver the corresponding batteries to the designated test station, and coordinating the adjustment of the test environment temperature and the configuration of the corresponding charge / discharge states and power parameters. The entire execution process unfolds along the rapid scheduling task sequence as the timeline, eliminating the need for complex optimization iterations. It can issue execution instructions in a timely manner under conditions of limited scheduling latency, ensuring that the system promotes each test task in an orderly manner according to priority. In this mode, even if a completely global optimum cannot be achieved, a relatively reasonable trade-off can be made between time constraints, temperature control costs, and power stability, meeting the battery testing requirements under scenarios with limited scheduling latency.

[0075] In one specific embodiment, this invention successfully controls peak power within a safe range and significantly smooths the power curve while slightly increasing energy consumption, thus achieving the goal of grid load optimization. An example is shown below: Table 1 Battery Test Task Parameter Setting Table

[0076] Please refer to Table 1. The battery test task parameter setting table provides the specific operating condition configurations for five battery test tasks T1 to T5, including: the target temperatures for each task are −20 ℃, 25 ℃, 60 ℃, -10 ℃, and 25 ℃, respectively; the test durations are 40 min, 20 min, 25 min, 35 min, and 15 min, respectively; the workstation positions are P1, P3, P2, P4, and P1, respectively; the temperature control stabilization times required to reach and stabilize the corresponding target temperatures are 3.2 min, 1.5 min, 2 min, 2.8 min, and 1.5 min, respectively; the charge / discharge states are alternated between discharge and charge, with T1 and T4 being discharge conditions and T2, T3, and T5 being charge conditions; the corresponding average powers are 5 W, 3 W, 4 W, 4.5 W, and 2.5 W, respectively, used to verify the performance of the battery test scheduling method based on multi-objective optimization under different temperature, duration, position, and power combinations in the experiment.

[0077] Table 2 Constant Parameter Setting Table

[0078] Please refer to Table 2, which shows the constant parameter settings used in the battery test scheduling method based on multi-objective optimization: the constant energy consumption per unit temperature difference, C_energy, is set to 1.92 kJ / °C to quantify the energy consumption during temperature switching; the maximum allowable power of the system, P_max, is set to 10 W to constrain the total power limit during the test; the weight coefficients α, β, and γ of energy consumption, time, and power in the multi-objective cost function are 0.62, 0.08, and 0.38, respectively; and the weights ω_e, ω_t, and ω_l of the time factor, temperature factor, and power factor used for fast scheduling priority scoring are set to 0.65, 0.35, and 0.5, respectively.

[0079] In this experimental scenario, the initial conditions were set as follows: the current temperature of the temperature control box is 25℃, and the initial position of the robotic arm is at the central workstation P0. A one-way movement time matrix from P0 to each test workstation P1 to P4 was provided to quantify the time consumption of subsequent scheduling schemes on the transport path. In comparison, the traditional FIFO scheduling executes in the order T1→T2→T3→T4→T5, causing the temperature path of the temperature control box to repeatedly switch between 25℃, -20℃, 60℃, and -10℃, resulting in numerous temperature switching times and high energy consumption. In contrast, the temperature-power collaborative optimization scheduling of this invention automatically generates the execution order T2→T5→T3→T4→T1 through a hybrid optimization algorithm, ensuring that the temperature control path starts at 25℃, maintains it at 25℃, and then sequentially adjusts to 60℃, -10℃, and -20℃, with only two substantial temperature switching events. Under this optimized sequence, the total energy consumption was 225.5 kJ, the total test time was 153.5 min, and the maximum instantaneous power was controlled at 8.0 kW within a safe range, indicating that the temperature control energy consumption was significantly reduced and the power peak was effectively suppressed while taking into account the test duration.

[0080] Table 3 Performance Index Comparison Table

[0081] Please refer to Table 3, which compares the performance indicators of different scheduling strategies: In terms of total energy consumption, the traditional FIFO scheduling consumes 441.6 kJ, while the temperature-optimized scheme of this invention reduces it to 220.8 kJ, and the temperature-power co-optimization scheme consumes 225.5 kJ, resulting in approximately half the energy consumption; the number of temperature switching times is reduced from 5 for FIFO to 2 for this invention; in terms of scheduling time, FIFO takes 20.6 min, while the temperature-optimized and temperature-power co-optimized schemes of this invention take 18.5 min and 19.1 min respectively, showing an overall improvement in scheduling efficiency; in terms of power characteristics, the maximum instantaneous power of FIFO is 12.5 kW and exceeds the limit, while the two schemes of this invention control it to 9.5 kW and 8.0 kW respectively, and the power variance is reduced from "high" to "medium" and "low", indicating that under the premise of slightly increasing energy consumption, the temperature-power co-optimization scheme further smooths the power curve and significantly improves grid friendliness.

[0082] Example 2 In addition, combined Figure 1 The battery test scheduling method based on multi-objective optimization driven by the embodiments of the present invention described herein can be implemented by test equipment. Figure 4 A schematic diagram of the hardware structure of the test equipment provided in an embodiment of the present invention is shown.

[0083] The testing equipment may include a processor and a memory storing computer program instructions.

[0084] Specifically, the processor may include a central processing unit (CPU), an application-specific integrated circuit (ASIC), or one or more integrated circuits that can be configured to implement embodiments of the present invention.

[0085] Memory may include non-persistent memory in computer-readable media, such as random access memory (RAM) and / or non-volatile memory, such as read-only memory (ROM) or flash RAM. Memory is an example of computer-readable media.

[0086] Computer-readable media include both permanent and non-permanent, removable and non-removable media, which can store information using any method or technology. Information can be computer-readable instructions, data structures, modules of programs, or other data. Examples of computer storage media include, but are not limited to, phase-change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technologies, CD-ROM, digital versatile optical disc (DVD) or other optical storage, magnetic tape, magnetic disk storage or other magnetic storage devices, or any other non-transferable medium that can be used to store information accessible by a computing device. As defined herein, computer-readable media does not include transient media, such as modulated communication signals and carrier waves.

[0087] The processor reads and executes computer program instructions stored in memory to implement any of the battery test scheduling methods based on multi-objective optimization in the above embodiments.

[0088] In one example, the test device may also include a communication interface and a bus. For example, Figure 4 As shown, the processor 401, memory 402, and communication interface 403 are connected through bus 410 and complete communication with each other.

[0089] The communication interface is mainly used to enable communication between various modules, devices, units and / or equipment in the embodiments of the present invention.

[0090] A bus, including hardware, software, or both, couples components of a test device together. For example, and not limitingly, a bus may include an Accelerated Graphics Port (AGP) or other graphics bus, an Enhanced Industry Standard Architecture (EISA) bus, a Front Side Bus (FSB), HyperTransport (HT) interconnect, an Industry Standard Architecture (ISA) bus, an Infinite Bandwidth Interconnect, a Low Pin Count (LPC) bus, a memory bus, a Microchannel Architecture (MCA) bus, a Peripheral Component Interconnect (PCI) bus, a PCI-Express (PCI-X) bus, a Serial Advanced Technology Attachment (SATA) bus, a Video Electronics Standards Association Local (VLB) bus, or other suitable buses, or combinations of two or more of these. Where appropriate, a bus may include one or more buses. While specific buses are described and illustrated in embodiments of the invention, the invention contemplates any suitable bus or interconnect.

[0091] In summary, the embodiments of the present invention provide a battery test scheduling method and device based on multi-objective optimization.

[0092] It should be clarified that the present invention is not limited to the specific configurations and processes described above and shown in the figures. For the sake of brevity, detailed descriptions of known methods are omitted here. In the above embodiments, several specific steps are described and shown as examples. However, the method process of the present invention is not limited to the specific steps described and shown. Those skilled in the art can make various changes, modifications, and additions, or change the order of steps, after understanding the spirit of the present invention.

[0093] Those skilled in the art will understand that embodiments of the present invention can be provided as methods, systems, or computer program products. Therefore, the present invention can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention can take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0094] This invention is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart illustrations and / or block diagrams. Figure 1 One or more processes and / or boxes Figure 1A device that provides the functions specified in one or more boxes.

[0095] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0096] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0097] It should also be noted that the exemplary embodiments mentioned in this invention describe methods or systems based on a series of steps or apparatus. However, this invention is not limited to the order of the steps described above; that is, the steps can be performed in the order mentioned in the embodiments, or in a different order, or several steps can be performed simultaneously.

[0098] The above description is merely a specific embodiment of the present invention. Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the specific working processes of the systems, modules, and units described above can be referred to the corresponding processes in the foregoing method embodiments, and will not be repeated here. It should be understood that the protection scope of the present invention is not limited thereto. Any person skilled in the art can easily conceive of various equivalent modifications or substitutions within the technical scope disclosed in the present invention, and these modifications or substitutions should all be covered within the protection scope of the present invention.

Claims

1. A battery test scheduling method based on multi-objective optimization, characterized in that, A battery test scheduling device based on multi-objective optimization is provided. The device includes: a controller; a test environment module connected to the controller for providing the required temperature conditions for battery testing; several battery test stations set within the test environment; a data acquisition module for acquiring data for battery test scheduling in real time and transmitting the acquired data to the controller; and a handling module connected to the controller for handling batteries between the loading / unloading area and the battery test stations. The method includes: In response to a preset scheduling trigger command, the data acquisition module acquires the real-time temperature in the battery test environment, the real-time charging and discharging parameters of the test battery, and the task attribute information of each task in the task sequence to be scheduled. Based on the preset decision variables, multi-objective cost function, and objective constraints, a multi-objective optimization model is constructed. The real-time temperature, real-time charging and discharging parameters, and task attribute information are input into the multi-objective optimization model to obtain the target task sequence; According to the target task sequence, the handling module and the test environment module are controlled to perform test scheduling for the batteries to be tested at each battery test station.

2. The battery test scheduling method based on multi-objective optimization as described in claim 1, characterized in that, The process of responding to a preset scheduling trigger command and acquiring real-time temperature in the battery testing environment, real-time charging and discharging parameters of the test battery, and task attribute information of each task in the task sequence to be scheduled through the data acquisition module includes: Based on preset scheduling trigger events, obtain scheduling trigger conditions, wherein the scheduling trigger events include the completion of the previous battery test task, the addition of a new battery test task, and the test timeout; The scheduling triggering conditions are monitored, and when the scheduling triggering conditions are met, the scheduling triggering instruction is obtained; According to the scheduling trigger command, the data acquisition module collects the test environment temperature, the charging and discharging current and the charging and discharging voltage of the test battery to obtain the real-time temperature and the real-time charging and discharging parameters. The attribute information of each task in the sequence of tasks to be scheduled is parsed to obtain the attribute information of each task.

3. The battery test scheduling method based on multi-objective optimization as described in claim 1, characterized in that, The step of constructing a multi-objective optimization model based on preset decision variables, multi-objective cost functions, and objective constraints includes: The execution order of tasks in the task sequence is used as the decision variable; Based on the preset constant parameters and initialization state parameters, the total temperature control energy consumption cost function, the total test time cost function, and the power penalty cost function are constructed respectively to obtain each target cost function; The multi-objective cost function is obtained by weighting each objective cost function according to the preset weight coefficients. The target constraints are determined based on preset constraints on the number of task executions, the continuity of the transport path, and the power. The multi-objective optimization model is constructed based on the decision variables, the multi-objective cost function, and the objective constraints.

4. The battery test scheduling method based on multi-objective optimization as described in claim 3, characterized in that, Based on preset constant parameters and initialization state parameters, the total temperature control energy consumption cost function, the total test time cost function, and the power penalty cost function are constructed respectively, resulting in the following target cost functions: Obtain the constant parameters and initialization state parameters, wherein the constant parameters include a constant energy consumption per unit temperature difference and a power threshold, and the initialization state parameters include the initial test environment temperature, the initial position of the transport module, and the initial total charging and discharging power; Based on the unit temperature difference energy consumption constant and the initial test environment temperature, combined with the target temperatures of adjacent tasks in the task sequence, the total temperature control energy consumption cost function is constructed. The total test time cost function is constructed based on the movement time of the transport module, the stability time of the test environment, and the test duration of the task. Based on the initial total charging and discharging power, the power threshold, and the total power curve during the execution of the task sequence, and combined with a preset penalty coefficient, the power penalty cost function is constructed. Based on the total temperature control energy consumption cost function, the total test time cost function, and the power penalty cost function, determine each of the target cost functions.

5. The battery test scheduling method based on multi-objective optimization as described in claim 4, characterized in that, The step of constructing the total temperature control energy consumption cost function based on the unit temperature difference energy consumption constant and the initial test environment temperature, combined with the target temperatures of adjacent tasks in the task sequence, includes: Based on the task execution order in the task sequence, obtain the first target temperature of the first task and the second target temperature of the adjacent task; Based on the first target temperature, the second target temperature, and the initial test environment temperature, multiple temperature differences are calculated. The total temperature control energy consumption cost function is obtained by integrating the temperature difference values ​​and the unit temperature difference energy consumption constant.

6. The battery test scheduling method based on multi-objective optimization as described in claim 4, characterized in that, The construction of the total test time cost function based on the movement time of the transport module, the stability time of the test environment, and the duration of the task test includes: Based on the initial position of the transport module and the position of the battery testing station corresponding to the first task in the task sequence, the target movement time required for the transport module is calculated. The total sequence movement time is calculated based on the battery test station positions corresponding to adjacent tasks in the task sequence and the preset movement time matrix. The target stabilization time is calculated based on the initial test environment temperature and the target temperature corresponding to the first task. The total sequence stabilization time is calculated based on the target temperatures corresponding to adjacent tasks in the task sequence. The test duration of each task in the task sequence is summed to obtain the test duration of the task. The total test time cost function is obtained by summing the target movement time, total sequence movement time, target stabilization time, total sequence stabilization time, and task test duration.

7. The battery test scheduling method based on multi-objective optimization as described in claim 4, characterized in that, The step of constructing the power penalty cost function based on the initial total charging and discharging power, the power threshold, and the total power curve during the execution of the task sequence, combined with a preset penalty coefficient, includes: Obtain the penalty coefficient, wherein the penalty coefficient includes a first penalty coefficient and a second penalty coefficient; Based on the initial total charging and discharging power, the charging and discharging state of each task in the task sequence, the average charging and discharging power, and the preset test duration, the curve of the total power changing with time during the execution of the task sequence is simulated and constructed to obtain the total power curve. The power over-limit penalty item is calculated based on the total power curve, the first penalty coefficient, and the power threshold. Calculate the power variance within a preset time interval based on the total power curve; The power fluctuation penalty term is calculated based on the power variance and the second penalty coefficient. The power over-limit penalty term and the power fluctuation penalty term are summed to obtain the power penalty cost function.

8. The battery test scheduling method based on multi-objective optimization as described in claim 3, characterized in that, The step of inputting the real-time temperature, real-time charge / discharge parameters, and task attribute information into the multi-objective optimization model to obtain the target task sequence includes: The total real-time charging and discharging power is calculated based on the real-time charging and discharging parameters. Based on the task attribute information, obtain the real-time target temperature, real-time test duration, real-time workstation location, real-time charging and discharging status, and real-time average charging and discharging power of each task in the task sequence to be scheduled. Based on the real-time total charging and discharging power, real-time target temperature, real-time test duration, real-time workstation position, real-time charging and discharging status, and real-time average charging and discharging power, the parameters of the multi-objective optimization function in the multi-objective optimization model are updated to obtain the updated multi-objective optimization model. Based on the updated multi-objective optimization model, the differential evolution algorithm is used to perform a global search on the solution space with the task execution order as the decision variable, and the multi-objective cost function values ​​of multiple candidate solutions representing different task execution orders are calculated and evaluated to obtain a candidate task sequence. The candidate task sequence is substituted into a preset integer programming solution model as the initial solution to optimize the task execution order under the target constraints, thereby obtaining the target task sequence.

9. The battery test scheduling method based on multi-objective optimization according to any one of claims 1-8, characterized in that, If the current battery test scenario is a preset scheduling latency-constrained scenario, the method further includes: Based on the waiting time of each task in the task sequence to be scheduled, the temperature difference between the current test environment temperature and the real-time target temperature of each task, and the deviation between the total system power predicted after each task is executed based on the real-time total charging and discharging power, real-time charging and discharging status and real-time average charging and discharging power and the preset ideal average power, the evaluation index values ​​used to characterize the time factor, temperature factor and power factor are determined respectively. Based on the preset evaluation weight coefficients, the values ​​of each evaluation index are weighted and calculated to obtain the priority score value corresponding to each task in the task sequence to be scheduled. Based on the priority score values, each task in the sequence of tasks to be scheduled is sorted, and target tasks are selected in order of priority score from high to low to generate a fast scheduling task sequence for the current scheduling cycle. According to the fast scheduling task sequence, the transport module and the test environment module are controlled to perform test scheduling on the batteries to be tested corresponding to each target task in the fast scheduling task sequence, so as to achieve battery testing under the constraint of scheduling delay.

10. A testing device, characterized in that, include: At least one processor, at least one memory, and computer program instructions stored in the memory, which, when executed by the processor, implement the method as described in any one of claims 1-9.

Images