Intelligent management method and system for energy storage cabinet

Abstract

This invention relates to the field of energy storage cabinet management technology, specifically to an intelligent management method and system for energy storage cabinets. The method includes the following steps: assessing the equipment's highest temperature and state of charge (SOC), obtaining deviation states, classifying the equipment into different load states to generate a division result, generating and obtaining the upper limit of the load-bearing capacity based on the division result, constructing a safety margin matrix using absolute power margin, calculating the total deficit and allocation ratio coefficient to obtain the target incremental power, superimposing to generate expected instructions, and adjusting and outputting execution instructions. In this invention, by coordinating the constraint conflicts between the overall instructions and local limitations, instantaneous reduction and repeated loading caused by the interaction between protection actions and follow-up logic are avoided. This solves the serious defect of equipment falling into reciprocating oscillations under continuous discharge conditions, curbs large power fluctuations to balance load-bearing capacity and operating boundaries, alleviates aging losses and effectively releases the system's dispatchable capacity, and ensures smooth response to instructions to stably complete dispatch tasks.

Description

Technical Field

[0001] This invention relates to the field of energy storage cabinet management technology, and in particular to an intelligent management method and system for energy storage cabinets. Background Technology

[0002] The field of energy storage cabinet management technology mainly involves the core aspects of state monitoring and operation scheduling of battery energy storage systems. This field systematically coordinates and regulates the physical environment of the energy storage and release process by collecting the voltage, current, and temperature values ​​of individual battery cells inside the energy storage cabinet, combined with the fan speed and air conditioning cooling capacity settings within the cabinet. Traditional intelligent management methods for energy storage cabinets focus on the execution of charging and discharging actions and thermal runaway early warning technologies for battery clusters within the cabinet. This typically involves continuously collecting physical quantities using thermocouple sensors attached to the battery pack casing and Hall effect current sensors connected in series with the power supply line. The instantaneous temperature and current values ​​are compared with fixed threshold values ​​pre-written into a storage chip. When the temperature or current value exceeds the set threshold, a high-level command is sent directly to the contactor to disconnect the battery main circuit, or a pulse-width modulation signal is sent to the cooling fan motor to increase the fan speed. Simultaneously, changes in battery terminal voltage and charging / discharging current are recorded according to a set time step.

[0003] In traditional management methods, during actual operation scenarios involving continuous full-power discharge, the power station experiences prolonged deep discharge, causing the cabinet to enter the critical zone. At this point, a single temperature-based hard protection logic is triggered, resulting in an instantaneous power reduction. The load reduction causes the temperature to drop, causing the protection judgment condition to disappear. To compensate for the shortfall, the system's power following logic will increase the cabinet power again. This rigid conflict between the single over-limit judgment action and the power following command causes the equipment to fall into a repetitive oscillation cycle of load reduction and recovery, resulting in significant fluctuations in the total output power of the power station. This makes it impossible to effectively complete the scheduling task and accelerates battery aging. Summary of the Invention

[0004] The purpose of this invention is to address the shortcomings of existing technologies by proposing an intelligent management method and system for energy storage cabinets.

[0005] To achieve the above objectives, the present invention adopts the following technical solution: an intelligent management method for energy storage cabinets, comprising the following steps: S1: Collect and process the basic operating status characteristics within the power plant control cycle, construct a basic input information matrix, extract dynamic polarization internal resistance based on the basic input information characteristics to calculate real-time heat generation power, combine the convective heat transfer coefficient to calculate the equipment heat dissipation power, and calculate the net heat accumulation rate based on the real-time heat generation power and the equipment heat dissipation power. S2: Calculate the ratio of the net heat accumulation rate to the set equivalent heat capacity parameter to obtain the predicted temperature rise rate, evaluate the highest single-unit temperature and the current state of charge to obtain the temperature deviation state and the power deviation state, and classify each of the energy storage cabinets into the normal load state, critical load state, limited recovery state or prohibited heavy load state according to the predicted temperature rise rate, the temperature deviation state and the power deviation state, and generate the operation state classification result. S3: Filter the first type of energy storage cabinets that belong to the critical load-bearing state according to the operation state division results. Generate the maximum allowable heat generation power based on the preset maximum allowable temperature rise rate of the first type of energy storage cabinet and the heat dissipation power of the equipment. Calculate the upper limit of the evaluated output power. Filter the second type of energy storage cabinets that belong to the limited recovery state according to the operation state division results. Extract the actual maintenance power of the second type of energy storage cabinet as the locked upper limit of output power. Establish cabinet hierarchical constraint mapping association to obtain the upper limit of the load-bearing output power. S4: Calculate the difference between the upper limit of the capacity to be supplied and the current actual output to obtain the absolute power margin; calculate the thermal state health factor based on the highest unit temperature and the current state of charge, and construct the station's remaining safety margin matrix based on the absolute power margin and the thermal state health factor; S5: Calculate the total power deficit and allocation ratio coefficient of the power station based on the remaining safety margin matrix of the entire station and the total power dispatching instruction of the power station, obtain the preliminary target incremental power, superimpose the preliminary target incremental power onto the current actual output to generate the expected power instruction, obtain the truncation limit instruction by selecting the smaller value between the expected power instruction and the upper limit of the load capacity, and adjust the power change rate of the truncation limit instruction using a preset slope limiter to obtain the target power execution instruction.

[0006] The present invention is improved in that the specific steps of S1 are as follows: S111: Collect basic operating status characteristics within the power station control cycle. The basic operating status characteristics include the highest single cell temperature of the battery, the current state of charge, the current charging and discharging current, the DC bus voltage, the speed and frequency of the cooling equipment, and the current actual output. Apply a sliding window filter to the basic operating status characteristics to calculate the truncated average value within the preset sampling window. Smooth the truncated average value according to the first-order low-pass filtering algorithm to construct the basic input information matrix. S112: Based on the relationship between the battery's historical operating temperature and internal resistance, a dynamic heat-sensitivity mapping matrix is ​​constructed to reflect the dynamic nonlinear mapping relationship between different states of charge and internal resistance. The current state of charge in the basic input information matrix is ​​called to match the internal resistance feature nodes in the dynamic heat-sensitivity mapping matrix to extract the dynamic polarization internal resistance. Based on the current charging and discharging current and the dynamic polarization internal resistance in the basic input information matrix, the product of the square parameter of the current charging and discharging current and the dynamic polarization internal resistance is calculated to obtain the real-time heat generation power. S113: Obtain the effective heat dissipation area of ​​the equipment and the set ambient temperature difference, map the energy storage cabinet as a heat accumulation model heating node, obtain the convective heat transfer coefficient of the heat accumulation model heating node according to the rotation speed frequency of the cooling equipment, calculate the product of the convective heat transfer coefficient, the effective heat dissipation area and the set ambient temperature difference to obtain the equipment heat dissipation power, and calculate the net heat accumulation rate according to the real-time heat generation power and the equipment heat dissipation power.

[0007] The present invention is improved in that the process of obtaining the convective heat transfer coefficient of the heating node in the heat accumulation model based on the rotational speed frequency matching of the cooling equipment is specifically as follows: By adjusting the cooling equipment to multiple fixed speed nodes under a preset standard test environment and collecting the surface heat flux data of the energy storage cabinet, a convective heat transfer mapping model is constructed through correlation calculation. Extract the rotational speed frequency of the cooling equipment and import the rotational speed frequency of the cooling equipment into the convective heat transfer mapping model; Interpolation is performed on the discrete data nodes within the convection heat transfer mapping model to extract the output heat transfer values ​​corresponding to the rotational speed frequency of the cooling equipment. The output heat transfer value is set as the convective heat transfer coefficient of the heating node in the heat accumulation model; The specific process for calculating the net heat accumulation rate is as follows: Calculate the power difference between the real-time heat generation power and the heat dissipation power of the equipment; The thermal conductivity of the internal support structure material of the energy storage cabinet is obtained and a preset reference thermal conductivity is set. A heat accumulation correction coefficient is set based on the ratio of the thermal conductivity of the internal support structure material of the energy storage cabinet to the preset reference thermal conductivity. Calculate the product of the power value difference and the heat accumulation correction factor, and use the product as the net heat accumulation rate.

[0008] The present invention is improved in that the specific steps of S2 are as follows: S211: Obtain the set equivalent heat capacity parameter of the battery module, calculate the ratio of the net heat accumulation rate to the set equivalent heat capacity parameter, and obtain the predicted temperature rise rate. S212: Call the basic input information matrix, extract the highest single-cell temperature and the current state of charge, obtain the preset temperature warning line, the preset forced load reduction line and the preset state of charge limit, compare the numerical deviation of the highest single-cell temperature with the preset temperature warning line and the preset forced load reduction line to obtain the temperature deviation state, and compare the numerical deviation of the current state of charge with the preset state of charge limit to obtain the charge deviation state. S213: Combine the predicted temperature rise rate, the temperature deviation state, and the power deviation state to determine the thermodynamic constraint boundary state of the energy storage cabinet. Based on the thermodynamic constraint boundary state, classify each of the energy storage cabinets into normal load-bearing state, critical load-bearing state, limited recovery state, or prohibited heavy load state, establish an operational quadrant distribution mapping association, and generate the operational state classification result of each of the energy storage cabinets.

[0009] The present invention is improved in that the specific steps of S3 are as follows: S311: Filter the first type of energy storage cabinet that belongs to the critical load state according to the operation state division result, obtain the preset maximum allowable temperature rise rate matched by the first type of energy storage cabinet, call the set equivalent heat capacity parameter and the equipment heat dissipation power, calculate the product of the set equivalent heat capacity parameter and the preset maximum allowable temperature rise rate, and add the equipment heat dissipation power to the product to generate the maximum allowable heat generation power. S312: Calculate the square root of the ratio between the maximum allowable heat generation power and the dynamic polarization internal resistance to obtain the maximum allowable discharge current corresponding to the first type of energy storage cabinet. Call the basic input information matrix to extract the DC-side bus voltage. Calculate the product of the maximum allowable discharge current and the DC-side bus voltage to obtain the upper limit of the evaluated output. S313: Filter the second type of energy storage cabinets that belong to the restricted recovery state according to the operation state division result, read the restricted constraint data packets reported by the battery management device corresponding to the second type of energy storage cabinet, extract the actual maintenance power recorded in the restricted constraint data packets, set the actual maintenance power as the locked output limit corresponding to the second type of energy storage cabinet, aggregate the evaluated output limit and the locked output limit corresponding to different quadrants, establish cabinet level constraint mapping association, and obtain the upper limit of the output that can be accepted.

[0010] The present invention is improved in that the process of obtaining the maximum allowable discharge current corresponding to the first type of energy storage cabinet is specifically as follows: Obtain the preset limit discharge current and stable operating current, and set the discharge safety correction coefficient based on the ratio of the limit discharge current to the stable operating current. Calculate the numerical ratio of the maximum allowable heat generation power to the dynamic polarization internal resistance; Perform a square root operation on the numerical ratio to extract the root value data; Calculate the product of the root value data and the discharge safety correction coefficient to determine the maximum allowable discharge current corresponding to the first type of energy storage cabinet; The process of obtaining the upper limit of the load capacity is as follows: Construct a mapping form that includes a hardware identification field and an output limit field; Extract the hardware code of the first type of energy storage cabinet and fill it into the hardware identifier field, and store the evaluated output limit into the corresponding output limit field; Extract the hardware code of the second type of energy storage cabinet and fill it into the hardware identifier field, and store the locked output limit into the corresponding output limit field; Extract the record values ​​of all the output limit fields in the mapping form to obtain the upper limit of the capacity to be accepted.

[0011] The present invention is improved in that the specific steps of S4 are as follows: S411: Call the basic input information matrix, extract the current actual output, calculate the difference between the upper limit of the load-bearing output and the current actual output, and obtain the absolute power margin; S412: Obtain a preset set limit temperature and a set temperature range; call the basic input information matrix to extract the highest single-unit temperature; calculate the temperature value deviation based on the set limit temperature and the highest single-unit temperature; calculate the ratio of the temperature value deviation to the set temperature range to obtain the temperature distance ratio; call the basic input information matrix to extract the current state of charge; obtain a preset state of charge limit; calculate the energy value deviation based on the current state of charge and the preset state of charge limit to obtain a set energy range; calculate the ratio of the energy value deviation to the set energy range to obtain the state of charge distance ratio; obtain a preset temperature margin weight and a preset state of charge margin weight; calculate a first product based on the preset temperature margin weight and the temperature distance ratio; calculate a second product based on the preset state of charge margin weight and the state of charge distance ratio; calculate the sum of the first product and the second product to generate a thermal state health factor. S413: Calculate the product of the absolute power margin and the thermal health factor to obtain the comprehensive power margin, obtain the cabinet identifier, aggregate the operating status division result, the upper limit of the capacity, the current actual output, the absolute power margin, the thermal health factor and the comprehensive power margin based on the cabinet identifier, establish a hierarchical data mapping structure, and generate the total station remaining safety margin matrix.

[0012] The present invention is improved in that the specific steps of S5 are as follows: S511: Obtain the total power dispatch instruction of the power station, call the operating status division result in the remaining safety margin matrix of the entire station, filter the available energy storage cabinets in the normal load-bearing state and the critical load-bearing state, extract the upper limit of the capacity to be carried and the comprehensive power margin corresponding to the available energy storage cabinets, calculate the sum for all the upper limits of the capacity to be carried to obtain the total safe output of the entire station, calculate the numerical difference between the total power dispatch instruction of the power station and the total safe output of the entire station to obtain the total power deficit of the power station, and calculate the sum for all the comprehensive power margins to obtain the total comprehensive margin of the entire station; S512: Calculate the ratio of the comprehensive power margin of a single available energy storage cabinet to the total comprehensive power margin of the entire station to obtain the allocation ratio coefficient; calculate the product of the total power deficit of the power station and the allocation ratio coefficient to obtain the preliminary target incremental power; extract the current actual output corresponding to the available energy storage cabinet; calculate the sum of the current actual output and the preliminary target incremental power to obtain the expected power command; select the minimum value between the expected power command and the upper limit of the acceptable output to obtain the truncation limit command; S513: Extract the thermal state health factor corresponding to the available energy storage cabinet, obtain a preset slope limiter, match the maximum power change rate within the preset slope limiter according to the thermal state health factor, and adjust the cutoff limit command according to the maximum power change rate using the preset slope limiter to obtain the target power execution command.

[0013] The present invention is improved in that the process of obtaining the truncation restriction instruction is specifically as follows: Compare the expected power command with the numerical value of the maximum output capacity that can be supplied; When it is determined that the expected power command is greater than the maximum acceptable output power, the maximum acceptable output power is assigned to the cut-off limit command; When it is determined that the expected power command is not greater than the maximum output capacity, the expected power command is assigned to the cut-off limit command; The process of adjusting the cutoff limit command according to the maximum power change rate using the preset slope limiter is specifically as follows: Obtain the historical output power of the available energy storage cabinet in the previous control cycle; The absolute value of the difference between the cutoff limit command and the historical output power is calculated as the power command change. Obtain the hardware response delay time of the energy storage converter and the set standard control cycle, and map and set the command buffer coefficient according to the ratio of the hardware response delay time of the energy storage converter to the set standard control cycle. The dynamic limiting gradient is obtained by multiplying the maximum power change rate by the command buffer coefficient. Compare the magnitude of the power command change with the value of the dynamic limiting gradient; When it is determined that the change in the power command is greater than the dynamic limiting gradient, the sum of the historical output power and the dynamic limiting gradient is calculated, and the sum is set as the target power execution command. When it is determined that the change in the power command is not greater than the dynamic limiting gradient, the truncation limiting command is used as the target power execution command.

[0014] An intelligent management system for energy storage cabinets, wherein the intelligent management system for energy storage cabinets is used to implement the above-mentioned intelligent management method for energy storage cabinets, the system comprising: The multi-dimensional feature analysis module collects and processes the basic operating status features within the power plant control cycle, constructs a basic input information matrix, extracts dynamic polarization internal resistance based on the basic input information features to calculate real-time heat generation power, calculates equipment heat dissipation power in combination with convective heat transfer coefficient, and calculates net heat accumulation rate based on the real-time heat generation power and the equipment heat dissipation power. The load condition assessment module calculates the ratio of the net heat accumulation rate to the set equivalent heat capacity parameter to obtain the predicted temperature rise rate, assesses the highest single-unit temperature and the current state of charge to obtain the temperature deviation state and the power deviation state, and classifies each of the energy storage cabinets into normal load state, critical load state, limited recovery state or prohibited heavy load state according to the predicted temperature rise rate, the temperature deviation state and the power deviation state, and generates the operation state classification result. The capacity boundary estimation module filters out the first type of energy storage cabinets that belong to the critical load-bearing state based on the operation state division results. It generates the maximum allowable heat generation power based on the preset maximum allowable temperature rise rate of the first type of energy storage cabinet and the heat dissipation power of the equipment, and calculates the upper limit of the estimated output. It then filters out the second type of energy storage cabinets that belong to the limited recovery state based on the operation state division results, extracts the actual maintenance power of the second type of energy storage cabinet as the locked upper limit of output, establishes cabinet hierarchical constraint mapping association, and obtains the upper limit of the load-bearing capacity. The whole-site resource aggregation module calculates the difference between the upper limit of the capacity to be carried and the current actual output to obtain the absolute power margin; calculates the thermal state health factor based on the highest unit temperature and the current state of charge; and constructs the whole-site remaining safety margin matrix based on the absolute power margin and the thermal state health factor. The flexible collaborative allocation module calculates the total power deficit and allocation ratio coefficient of the power station based on the remaining safety margin matrix of the entire station and the total power dispatching command of the power station, obtains the preliminary target incremental power, superimposes the preliminary target incremental power onto the current actual output to generate the expected power command, obtains the truncation limit command by selecting the smaller value between the expected power command and the upper limit of the load capacity, and adjusts the power change rate of the truncation limit command using a preset slope limiter to obtain the target power execution command.

[0015] Compared with the prior art, the advantages and positive effects of the present invention are as follows: In this invention, the deviation state is obtained by calculating and predicting the temperature rise rate and evaluating the maximum temperature and state of charge. The equipment is classified into different load states to generate classification results. Based on the critical load and limited recovery state, the upper limit of output is evaluated and locked. The constraint mapping is established to obtain the upper limit of the load that can be carried. The absolute power margin and thermal health factor are calculated in combination with the actual output. The remaining safety margin matrix is ​​constructed and the total power deficit and amortization coefficient are calculated. The target incremental power is obtained and superimposed to generate the expected command. After truncation and slope adjustment, the target execution command is output. This effectively coordinates multiple constraint conflicts to avoid repeated loading and oscillation caused by hard protection, takes into account the safe operation boundary and extends the battery life to stably complete the scheduling task. Attached Figure Description

[0016] Figure 1 This is the main flowchart of the intelligent management of the energy storage cabinet of the present invention; Figure 2 This is a flowchart illustrating the process of generating the heat accumulation rate in this invention. Figure 3 This is a flowchart illustrating the operational state division and output upper limit determination of the present invention. Figure 4 This is a flowchart of the safety margin allocation and power command generation process of the present invention. Detailed Implementation

[0017] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the invention.

[0018] Please see Figures 1 to 4 This embodiment provides an intelligent management method for energy storage cabinets. In practical applications, such as when multiple energy storage cabinets are connected to the power station's energy management control link within an energy storage power station, and the cabinet's battery management equipment, energy storage converter, cooling equipment controller, and power station dispatch control node continuously exchange operating data, the battery thermal state, state of charge, output state, and cooling state of the energy storage cabinets are synchronously processed within the same control cycle, including the following steps: S1: Collect and process basic operating status characteristics within the power plant control cycle to construct a basic input information matrix. This matrix is ​​a data object created for each individual energy storage cabinet within the control cycle. It contains fields for battery temperature, state of charge, current, voltage, cooling equipment operation, and actual output, and includes associated cabinet identifiers, data source identifiers, acquisition cycle identifiers, and validity identifiers. This matrix is ​​generated after receiving raw data from the acquisition interface and enters the thermal state calculation stage. It is used to extract dynamic polarization resistance and form real-time heat generation power, while combining this with the convective heat transfer coefficient of the cooling equipment to form the equipment's heat dissipation power. The heat generation and heat dissipation data are then converted into a net heat accumulation rate, serving as input for subsequent temperature rise rate prediction and operating state classification.

[0019] S111: Collects basic operating status characteristics within the power station control cycle. These characteristics include the highest single-cell temperature, current state of charge (SBC), current charge / discharge current, DC bus voltage, cooling equipment rotation speed / frequency, and current actual power output. The highest single-cell temperature is generated by the cabinet battery management device selecting the highest temperature field within the current control cycle from the single-cell temperature acquisition channel. The current SBC is provided by the SBC estimation channel of the battery management device. The current charge / discharge current is provided by the DC current sampling channel. The DC bus voltage is provided by the DC sampling channel of the energy storage converter. The cooling equipment rotation speed / frequency is provided by the cooling equipment controller operation feedback channel. The current actual power output is provided by the power feedback channel of the energy storage converter.

[0020] A sliding window filter is applied to calculate the truncated average value within a preset sampling window based on the characteristics of the basic operating state. The sliding window filter is a sequential processing rule that caches the same field according to the acquisition time sequence. The cache location is located in the data preprocessing cache area of ​​the cabinet control node. The truncated average value is a smoothed field formed after removing distorted samples, duplicate samples, and out-of-bounds samples within the preset sampling window. The preset sampling window is issued by the controller parameter configuration area. The window boundary is derived from the power plant control cycle configuration and sensor refresh strategy. The update condition is the advancement of the control cycle or the update of the control strategy version. Low-pass smoothing is performed on the truncated average value. The low-pass smoothing process takes the retained value of the previous valid cycle and the current truncated average value as input and outputs the smoothed basic operating state field. When the equipment starts up and the first control cycle lacks historical retained values, the current sampled field with validity verification is used as the initial smoothing benchmark. When the historical window is not yet complete, the window processing range is supplemented by the already obtained valid sampled fields. When the sampled field is missing or communication timeout occurs, the previous valid field is used and a de-confidence status identifier is added for identification in the subsequent state division stage.

[0021] S112: Based on the historical operating temperature and internal resistance variation of the battery, a dynamic heat sensitivity mapping matrix is ​​constructed to reflect the dynamic nonlinear mapping relationship between different states of charge and internal resistance. The dynamic heat sensitivity mapping matrix is ​​a data mapping object stored in the parameter area of ​​the cabinet control node. Its fields include state of charge segment identifier, temperature segment identifier, internal resistance characteristic node, node source identifier, and valid version identifier. This mapping matrix is ​​generated jointly from battery factory test data, power station historical operating records, and maintenance calibration records, and is synchronously refreshed when parameter configuration is updated.

[0022] The current state of charge (SOC) from the basic input information matrix is ​​retrieved, and the internal resistance feature node is matched in the dynamic heat sensitivity mapping matrix. If the SOC falls into an existing node, the corresponding dynamic polarization internal resistance is directly read. If the SOC lies between adjacent nodes, interpolation is performed based on the order of adjacent nodes in the mapping matrix to obtain the dynamic polarization internal resistance corresponding to the current rack operating state. The dynamic polarization internal resistance is an intermediate field reflecting the battery's heat sensitivity due to polarization under the current SOC and thermal conditions, and is output to the real-time heat generation power generation stage. If no valid node is matched, the backup mapping range for the same battery model is read, and a mapping degradation flag is set. If the backup mapping is still unavailable, the rack is marked as a candidate for a prohibited overload state, and subsequent steps will no longer prioritize it for accepting incremental power.

[0023] Real-time heat generation power is generated based on the current charge / discharge current and dynamic polarization resistance within the basic input information matrix. This real-time heat generation power is a power field on the heat generation side, formed by the combined effect of the thermal effect intensity caused by the current current amplitude and the dynamic polarization resistance, and is generated within the thermal state calculation buffer. The direction of the current charge / discharge current is used to identify the charging or discharging condition, the current amplitude is used for heat sensitivity processing, and the dynamic polarization resistance is used to limit the internal resistance heat contribution under this current condition. This real-time heat generation power is output to the device heat dissipation power comparison stage and serves as the heat generation input for generating the net heat accumulation rate.

[0024] S113: Obtain the effective heat dissipation area of ​​the equipment and the set ambient temperature difference, and map the energy storage cabinet as a heat accumulation model heating node. The heat accumulation model heating node is a virtual node object for a single energy storage cabinet during the thermal state calculation phase, carrying the cabinet identifier, heat generation side input, heat dissipation side input, material correction field, and net heat accumulation field. The effective heat dissipation area of ​​the equipment is derived from the cabinet structural parameter configuration, and the set ambient temperature difference is derived from the deviation between the cabinet operating environment monitoring field and the thermal management control target field. Both are processed by the heat dissipation side rules and, together with the convective heat transfer coefficient, generate the equipment's heat dissipation power.

[0025] The convective heat transfer coefficient of the heating node in the heat accumulation model is obtained by matching the cooling equipment's rotational speed frequency. The convective heat transfer coefficient is a lookup table field representing the cooling equipment's heat dissipation capacity on the cabinet surface and internal heat exchange channels under the current rotational speed frequency feedback. This field is generated through correlation calculation between the cooling equipment's fixed rotational speed node and the cabinet surface heat flux data under a preset standard test environment and stored in the convective heat transfer mapping model. The convective heat transfer mapping model is a mapping data table between discrete rotational speed nodes and heat transfer output values. The input is the cooling equipment's rotational speed frequency, and the output is the heat transfer value under the current cooling state. After the cooling equipment's rotational speed frequency enters this mapping model, the field validity is first verified, and then interpolation is performed between discrete data nodes to output the heat transfer value corresponding to the current rotational speed feedback. This heat transfer value is then set as the convective heat transfer coefficient of the heating node in the heat accumulation model. When rotational speed feedback is missing, the previous valid rotational speed field of the cooling controller is read and a cooling status confirmation flag is added; when the rotational speed feedback exceeds the boundary of the mapping model, the heat transfer value corresponding to the boundary node is used and a maintenance record is triggered.

[0026] The heat dissipation power of the equipment is generated based on the convective heat transfer coefficient, effective heat dissipation area, and set ambient temperature difference. The equipment heat dissipation power is a heat dissipation-side power field formed under the combined constraints of the current cabinet's cooling equipment operating status, structural heat dissipation capacity, and ambient temperature difference, and is output to the net heat accumulation rate generation stage. When calculating the net heat accumulation rate, the power difference between the real-time heat generation power and the equipment heat dissipation power within the same control cycle is first compared to form a heat increase / decrease direction field. Then, the thermal conductivity of the cabinet's internal support structure material and the reference thermal conductivity are obtained. These two values ​​are derived from the cabinet structure parameter table and the material reference configuration table, and are mapped to a heat accumulation correction coefficient through material thermal conductivity matching rules. The heat accumulation correction coefficient is used to correct the influence of the cabinet's internal structure on heat retention or diffusion. The corrected heat increase / decrease result is written into the net heat accumulation rate field. This field carries the cabinet identifier, control cycle identifier, and material correction status, and is transmitted to the predicted temperature rise rate processing stage of S2.

[0027] S2: Calculate the ratio of net heat accumulation rate to the set equivalent heat capacity parameter to obtain the predicted temperature rise rate. Evaluate the highest single-cell temperature and the current state of charge to obtain temperature deviation and charge deviation. Based on the predicted temperature rise rate, temperature deviation, and charge deviation, classify each energy storage cabinet into normal operating state, critical operating state, limited recovery state, or prohibited heavy load state, generating an operating state classification result. The set equivalent heat capacity parameter is a thermal inertia configuration field formed by the cabinet's battery modules and adjacent thermal structures, stored in the cabinet's thermal parameter configuration area, derived from battery module structural parameters, thermal management calibration records, and operation and maintenance configurations. The operating state classification result is a set of state identifiers output to a single energy storage cabinet. Subsequently, S3 filters cabinets corresponding to the critical operating state and limited recovery state, and S5 filters available energy storage cabinets.

[0028] S211: Obtain the set equivalent thermal capacity parameters of the battery module, and perform thermal inertia matching between the net heat accumulation rate field and the set equivalent thermal capacity parameter field to obtain the predicted temperature rise rate. The predicted temperature rise rate is an intermediate state field representing the impact of the current control cycle's heat accumulation trend on the cabinet battery temperature change. It originates from the net heat accumulation rate output by S1 and the equivalent thermal capacity parameters in the thermal parameter configuration area. When the net heat accumulation rate shows that the heat generation side is higher than the heat dissipation side, the predicted temperature rise rate is marked as a heating trend; when the net heat accumulation rate shows that the heat dissipation side covers the heat generation side, the predicted temperature rise rate is marked as a cooling or maintaining trend. This field is written to the operating status determination cache and, together with the temperature deviation status and the charge deviation status, enters the thermodynamic constraint boundary determination.

[0029] S212: Call the basic input information matrix to extract the highest single-cell temperature and current state of charge, and obtain the preset temperature warning line, preset forced load reduction line, and preset state of charge limit. The preset temperature warning line and preset forced load reduction line are stored in the thermal management rule table, derived from battery safety boundaries, cabinet thermal management strategies, and operation and maintenance configurations; the preset state of charge limit is stored in the energy management rule table, derived from battery operation protection strategies and power plant scheduling strategies. The above fields have version identifiers and activation status identifiers, and are synchronized from the power plant scheduling control node to the cabinet control node after rule updates.

[0030] The temperature deviation status is obtained by comparing the deviation between the highest single-cell temperature and the preset temperature warning line and the preset forced load reduction line. The temperature deviation status includes status indicators such as temperature safety, temperature approaching warning, temperature entering the load reduction boundary, and temperature exceeding the limit candidate. The identification results are written to the operation status judgment cache. The charge deviation status is obtained by comparing the deviation between the current state of charge and the preset state of charge limit. The charge deviation status includes status indicators such as charge available, charge approaching the lower limit, and charge limited candidate. When the highest single-cell temperature field is missing, the risk side field of other valid temperature samples in the same cabinet is read and the temperature sampling is marked as abnormal; when the state of charge field is missing, the previous valid state of charge of the battery management device is read and the charge status is marked as pending confirmation. The above abnormal identifications enter S213 along with the temperature deviation status and the charge deviation status.

[0031] S213: Determine the thermodynamic constraint boundary state of the energy storage cabinet by combining the predicted temperature rise rate, temperature deviation state, and power deviation state. The thermodynamic constraint boundary state describes whether the cabinet can continue to bear power output under the current heat accumulation trend, temperature safety boundary, and power safety boundary. The determination order is as follows: first, identify candidates for temperature exceeding the limit and candidates for power limitation; then, identify cabinets whose temperature is close to the warning level and whose predicted temperature rise rate is on the rising trend; finally, identify cabinets whose temperature is safe, whose power is available, and whose predicted temperature rise rate has not entered the direction of rising temperature risk. When multiple state conditions exist simultaneously, the risk-side priority rule is executed, and the state identifier with the higher degree of restriction is output first. The state identifier is written into the operation quadrant distribution mapping association.

[0032] Based on the thermodynamic constraint boundary states, each energy storage cabinet is categorized into normal operating state, critical operating state, limited recovery state, or prohibited overload state. The normal operating state identifies cabinets where both temperature and power deviations are within acceptable limits and the predicted temperature rise trend has not triggered limitations. The critical operating state identifies cabinets that can still handle some output but whose thermal state is close to the limit boundary. The limited recovery state identifies cabinets that need to maintain or reduce output to restore their thermal or power state. The prohibited overload state identifies cabinets that do not participate in incremental output allocation. The operating state classification results carry cabinet identifiers, state identifiers, state source fields, and anomaly identifiers, and are output to S3 and S4.

[0033] S3: First-class energy storage cabinets, whose operational status classification results correspond to the critical load-bearing state, are selected. Based on the preset maximum allowable temperature rise rate and equipment heat dissipation power of the first-class energy storage cabinets, the maximum allowable heat generation power is generated, and the estimated output limit is calculated. Second-class energy storage cabinets, whose operational status classification results correspond to the constrained recovery state, are selected. The actual maintenance power of the second-class energy storage cabinets is extracted as the locked output limit. A cabinet-level constraint mapping relationship is established to obtain the maximum acceptable output limit. The first-class energy storage cabinets are the set of cabinets that still have constrained incremental load-bearing capacity in the S2 state classification. The second-class energy storage cabinets are the set of cabinets that need to participate in the control closed loop with maintenance power. The maximum acceptable output limit is the output boundary field jointly invoked by the subsequent safety margin matrix and power scheduling constraints.

[0034] S311: Filter the energy storage cabinets whose operating status classification results belong to the first type of critical load-bearing state, and obtain the preset maximum allowable temperature rise rate matched to the first type of energy storage cabinet. The preset maximum allowable temperature rise rate is a configurable field in the thermal management rule table, derived from the battery thermal safety boundary, cooling capacity calibration, and power plant operation strategy. Field updates are managed by parameter version control rules. Call the set equivalent heat capacity parameters and equipment heat dissipation power, perform thermal inertia matching between the allowable thermal change capacity corresponding to the preset maximum allowable temperature rise rate and the equivalent heat capacity parameters, and add the current heat dissipation capacity that can be removed to generate the maximum allowable heat generation power. The maximum allowable heat generation power is the upper limit field of the heat generation side for the critical load-bearing state cabinet under the condition of not further expanding the thermal risk boundary, and is output to the maximum allowable discharge current generation stage.

[0035] S312: Generate the maximum allowable discharge current corresponding to the first type of energy storage cabinet based on the maximum allowable heat generation power and dynamic polarization internal resistance. The generation process first obtains the preset set limit discharge current and stable operating current, which are fields in the protection configuration of the energy storage converter and battery management equipment, derived from equipment factory limits, battery operation strategies, and power station control strategies; then, based on the correspondence between the two types of current fields in the rule table, a discharge safety correction coefficient is set. The discharge safety correction coefficient is set as a dimensionless correction field for controlling the discharge boundary of the cabinet under critical load conditions, stored in the power constraint parameter area, and updated with the equipment protection strategy.

[0036] After the maximum permissible heat generation power and dynamic polarization internal resistance are incorporated into the current boundary processing rules, a matching result between heat generation capacity and internal resistance sensitivity is first formed. Then, root value extraction processing is performed on the matching result to obtain root value data. The root value data and the discharge safety correction coefficient together form the maximum permissible discharge current. The DC-side bus voltage is extracted by calling the basic input information matrix. The maximum permissible discharge current and the DC-side bus voltage are matched against the power boundary under the same cabinet identifier to obtain the estimated output upper limit. The estimated output upper limit is written into the cabinet level constraint mapping association. Subsequently, S4 calculates the absolute power margin, and S5 performs truncation limitation.

[0037] S313: Filter the second-class energy storage cabinets whose operating status classification results correspond to the restricted recovery state, and read the restricted constraint data packets reported by the battery management device corresponding to the second-class energy storage cabinet. The restricted constraint data packet is a data object output by the battery management device after thermal protection, power protection, or internal operation protection is triggered. Fields include cabinet hardware code, constraint cause identifier, actual maintenance power, constraint validity status, and reporting cycle identifier. Extract the actual maintenance power recorded in the restricted constraint data packet and set this actual maintenance power as the locked output limit corresponding to the second-class energy storage cabinet. If the actual maintenance power field is missing or the constraint validity status fails, transfer the second-class energy storage cabinet to the prohibited overload state candidate and report the constraint data anomaly identifier to the power station dispatch control node.

[0038] Construct a mapping form containing a hardware identifier field and an output limit field. The mapping form is a storage object associated with rack-level constraint mapping. The hardware identifier field is used to write the hardware code of the energy storage rack, and the output limit field is used to write the estimated or locked output limit. Extract the hardware code of the first type of energy storage rack and fill it into the hardware identifier field, and store the estimated output limit in the corresponding output limit field; extract the hardware code of the second type of energy storage rack and fill it into the hardware identifier field, and store the locked output limit in the corresponding output limit field. Aggregate the estimated and locked output limits corresponding to different quadrants, extract the record values ​​of all output limit fields in the mapping form, and obtain the maximum acceptable output limit. The maximum acceptable output limit is output to S4 according to the rack identifier, and a power command cutoff boundary is provided to S5.

[0039] S4: Calculate the absolute power margin by the difference between the maximum capacity output and the current actual output. Calculate the thermal state health factor based on the highest unit temperature and the current state of charge. Construct a station-wide remaining safety margin matrix based on the absolute power margin and the thermal state health factor. The station-wide remaining safety margin matrix is ​​a data aggregation object of cabinet safety resources at the power station level. Its fields include cabinet identifier, operating status classification result, maximum capacity output, current actual output, absolute power margin, thermal state health factor, and comprehensive power margin. This matrix is ​​generated by the cabinet control node or the power station scheduling control node and serves as the basic input for S5 power allocation and slope limitation processing.

[0040] S411: Call the basic input information matrix to extract the current actual output. Identify the difference between the upper limit of the available output and the current actual output to obtain the absolute power margin. The absolute power margin is the remaining power space field between the current actual output of a single energy storage cabinet and the boundary of its available output. It originates from the upper limit of available output output output by S3 and the current actual output collected by S1. When the feedback of the current actual output is lagging, prioritize reading the previous valid output power confirmed by the energy storage converter and add a power feedback pending confirmation flag. When the upper limit of available output is missing, mark the absolute power margin as unavailable and exclude the cabinet from incremental capacity acceptance during S5 allocation.

[0041] S412: Obtain the preset set limit temperature and set temperature range; extract the highest single-cell temperature by calling the basic input information matrix; and obtain the temperature distance ratio based on the distance relationship between the set limit temperature and the highest single-cell temperature. The set limit temperature and set temperature range are configuration fields in the thermal management rule table, derived from the battery safety boundary and cabinet thermal management strategy. The temperature distance ratio is a dimensionless state field describing the distance of the highest single-cell temperature from the limit boundary. The current state of charge (SBC) is extracted by calling the basic input information matrix; the preset SBC limit is obtained; the energy deviation is obtained based on the distance relationship between the current SBC and the preset SBC limit; and the SBC distance ratio is obtained by combining the set energy range. The set energy range is a configuration field in the energy management rule table, derived from the battery available energy control strategy and scheduling operation boundary.

[0042] The system retrieves preset temperature margin weights and preset state-of-charge (SOC) margin weights, both configurable fields in the safety margin rule table. These weights originate from the power plant's operational strategy's priority configuration for thermal and electrical safety and are stored in the scheduling control parameter area. A temperature-side contribution result is generated based on the preset temperature margin weights and the temperature-distance ratio, while an electrical-side contribution result is generated based on the preset SOC margin weights and the SOC distance ratio. These two contributions are then combined into a thermal state health factor. The thermal state health factor is a dimensionless field describing the degree of power variation that a cabinet can handle under the combined constraints of temperature and SOC margins, and is output to S413 and S513. If any distance ratio field cannot be generated due to sampling anomalies, the corresponding anomaly flag is read, and the availability status of the health factor is reduced to prevent abnormal cabinets from receiving excessively high priority in subsequent allocation.

[0043] S413: A comprehensive power margin is obtained by comprehensively matching the absolute power margin and the thermal health factor. The comprehensive power margin is the rack-allocated capacity field that simultaneously considers the remaining space at the output boundary, the temperature safety distance, and the power safety distance. The rack identifier is obtained, and based on the rack identifier, the operating status division results, the upper limit of the capacity to be handled, the current actual output, the absolute power margin, the thermal health factor, and the comprehensive power margin are aggregated to establish a hierarchical data mapping structure and generate a total station remaining safety margin matrix. The hierarchical data mapping structure uses the station identifier as the upper-level index and the rack identifier as the lower-level index. Each field carries the source step identifier, update cycle identifier, and valid status identifier. After the matrix is ​​generated, it is written to the safety margin cache of the station scheduling and control node and provided to S5 to calculate the station's total power deficit, the allocation ratio coefficient, and the target power execution instruction.

[0044] S5: Based on the total remaining safety margin matrix of the entire station and the total power dispatch command of the power station, calculate the total power deficit and allocation ratio coefficient of the power station to obtain the preliminary target incremental power. This preliminary target incremental power is then superimposed on the current actual output to generate the expected power command. The smaller value between the expected power command and the upper limit of the available output is selected to obtain the truncation limit command. A preset slope limiter is used to adjust the power change rate of the truncation limit command to obtain the target power execution command. The total power dispatch command of the power station is issued by the power station energy management system and undergoes consistency verification with the total remaining safety margin matrix after entering the dispatch control node. The target power execution command is issued by the dispatch control node to the energy storage converter and fed back to the cabinet battery management equipment and cooling equipment controller to form a closed-loop record.

[0045] S511: Obtain the total power dispatch instruction for the power plant, call the operating status division results within the remaining safety margin matrix of the entire plant, and filter available energy storage cabinets in normal and critical load states. Available energy storage cabinets are the set of cabinets allowed to participate in total power allocation within the current control cycle; cabinets in restricted recovery state and prohibited overload state do not participate in incremental power assignment. Extract the upper limit of the output capacity and the comprehensive power margin corresponding to the available energy storage cabinets, summarize all the upper limits of the output capacity to obtain the total safe output capacity of the entire plant. The total safe output capacity of the entire plant is the output boundary field that can be issued by the entire plant under safety constraints.

[0046] The total power dispatch command of the power station is compared with the total safe output of the entire station to identify the total power deficit. The total power deficit represents the difference between the dispatch command and the current safe carrying capacity, and is then processed for incremental allocation. All comprehensive power margins are summarized to obtain the total comprehensive margin of the entire station. If the total comprehensive margin of the entire station is unavailable, the dispatch control node marks the total power deficit of the power station as pending load reduction and reports the insufficient safety margin status to the superior energy management system; if the comprehensive power margin field of some available energy storage cabinets is abnormal, only the abnormal cabinets are excluded and the remaining valid cabinets are retained for subsequent allocation.

[0047] S512: Calculate the allocation relationship between the comprehensive power margin of a single available energy storage cabinet and the total comprehensive margin of the entire station, obtaining the allocation ratio coefficient. The allocation ratio coefficient is the share of a single available energy storage cabinet in the remaining safety margin of the entire station, derived from the comprehensive power margin generated in S413 and the total comprehensive margin of the entire station generated in S511. The total power deficit of the power station and the allocation ratio coefficient together form the preliminary target incremental power, which is written into the cabinet power allocation cache. Extract the current actual output corresponding to the available energy storage cabinet, and superimpose the current actual output with the preliminary target incremental power to generate the expected power command. The expected power command is the target power field before the single-unit upper limit truncation and power change rate constraints.

[0048] Upon receiving a cutoff limit command, the expected power command is compared with the maximum acceptable output capacity. If the expected power command is higher than the maximum acceptable output capacity, the maximum acceptable output capacity is assigned to the cutoff limit command; if the expected power command is not higher than the maximum acceptable output capacity, the expected power command is assigned to the cutoff limit command. The cutoff limit command carries a cutoff source identifier, which is used to distinguish between cutoff triggered by the safe output boundary and direct issuance candidate states that have not triggered cutoff. This command enters the slope limiting processing of S513 to prevent single-cycle power changes from exceeding the dynamic response boundaries allowed by the thermal state of the energy storage converter and battery.

[0049] S513: Extract the thermal state health factor corresponding to the available energy storage cabinet and obtain the preset slope limiter. The preset slope limiter is a power change rate control rule object stored in the scheduling control parameter area. Its fields include a health level identifier, a maximum power change rate field, a hardware response correction field, and an enabled status field. Match the maximum power change rate within the preset slope limiter to the thermal state health factor. The closer the health factor is to an insufficient safety margin state, the more restricted the corresponding maximum power change rate; when the health factor is in a sufficient safety margin state, the corresponding power change constraint maintains the normal scheduling boundary. If matching fails, the default risk level in the rule table is used, with an additional slope parameter confirmation flag.

[0050] The process of adjusting the cutoff limit command according to the maximum power change rate using a preset slope limiter involves first obtaining the historical output power of available energy storage cabinets in the previous control cycle. This historical output power originates from the energy storage converter's confirmation feedback field and is stored in the power execution history cache. The change in the cutoff limit command is then identified by comparing it with the historical output power to obtain the power command change. Next, the hardware response delay time of the energy storage converter and the set standard control cycle are obtained; these are the energy storage converter control parameters and the power plant scheduling cycle configuration fields. Based on their matching relationship in the hardware response rule table, a command buffer coefficient is generated. This command buffer coefficient reflects the buffering requirement between the power command issuance and the actual hardware response and is output to the dynamic limiting gradient generation stage.

[0051] The maximum power change rate and the command buffer coefficient are used to generate a dynamic limiting gradient. The dynamic limiting gradient is the boundary field that allows power command changes in the current control cycle. The magnitude of the power command change and the dynamic limiting gradient are compared. When the power command change exceeds the dynamic limiting gradient, the output power is adjusted along the historical output power towards the truncation limit command direction according to the dynamic limiting gradient, and the adjusted power is set as the target power execution command. When the power command change does not exceed the dynamic limiting gradient, the truncation limit command is used as the target power execution command. The target power execution command is sent to the energy storage converter for execution, and the execution confirmation result is written back to the input field of the next control cycle in the basic input information matrix. If no execution confirmation is received, the previous valid execution command is retained and a control confirmation anomaly flag is generated for subsequent control cycles to participate in state division and power allocation.

[0052] The intelligent management system for energy storage cabinets in the same embodiment carries out the operation of the aforementioned method. This system includes a multi-dimensional feature analysis module, a load-bearing condition assessment module, a capacity boundary estimation module, a station-wide resource aggregation module, and a flexible collaborative allocation module. The multi-dimensional feature analysis module connects to the data interfaces of the battery management device, energy storage converter, and cooling equipment controller, and carries out the basic operating state feature acquisition, filtering, dynamic polarization resistance extraction, real-time heat generation, equipment heat dissipation power generation, and net heat accumulation rate output in S1. The load-bearing condition assessment module receives the net heat accumulation rate and basic input information matrix output by the multi-dimensional feature analysis module, and carries out the predicted temperature rise rate generation, temperature deviation state generation, power deviation state generation, and operating state division result output in S2. The capacity boundary estimation module receives the operating state division result, equipment heat dissipation power, dynamic polarization resistance, and DC-side bus voltage, and carries out the first type of energy storage cabinet screening, second type of energy storage cabinet screening, output limit evaluation generation, output limit locking generation, and capacity-bearing output limit output in S3. The station-wide resource aggregation module receives the maximum capacity output and the current actual output, and carries out the generation of absolute power margin, thermal health factor, comprehensive power margin, and storage of the station-wide remaining safety margin matrix in S4. The flexible collaborative allocation module receives the station-wide remaining safety margin matrix and the total power dispatching command of the power station, and carries out the identification of the total power deficit of the power station, generation of the allocation ratio coefficient, generation of the preliminary target incremental power, generation of the expected power command, generation of the truncation limit command, adjustment of the slope limit, and output of the target power execution command in S5.

[0053] Within the system, each module maintains a data correspondence through rack identifiers, control cycle identifiers, and field validity status identifiers. Before entering the load condition assessment module, the data output by the multi-dimensional feature parsing module undergoes field integrity verification. Before entering the capacity boundary estimation module, the status identifiers output by the load condition assessment module undergo status version verification. Before entering the full-site resource aggregation module, the output power boundary from the capacity boundary estimation module undergoes rack hardware code verification. Before entering the flexible collaborative allocation module, the matrix output by the full-site resource aggregation module undergoes valid row verification. After issuing the target power execution command, the flexible collaborative allocation module writes the execution confirmation status, control confirmation anomaly identifier, and power execution history to the scheduling log cache. The scheduling log cache provides traceability access according to access permissions, and the traceability content includes data source category, processing stage identifier, status identifier, anomaly feedback path, and execution confirmation status. The aforementioned module load relationships enable data acquisition, status assessment, boundary estimation, safety margin aggregation, and power collaborative allocation in the aforementioned method to form a closed loop within the same energy storage power station control link.

[0054] The above are merely preferred embodiments of the present invention and are not intended to limit the present invention in any other way. Any person skilled in the art may make changes or modifications to the above-disclosed technical content to create equivalent embodiments that can be applied to other fields. However, any simple modifications, equivalent changes, and modifications made to the above embodiments based on the technical essence of the present invention without departing from the scope of the present invention shall still fall within the protection scope of the present invention.

Claims

1. An intelligent management method for energy storage cabinets, characterized in that, Includes the following steps: S1: Collect and process the basic operating status characteristics within the power plant control cycle, construct a basic input information matrix, extract dynamic polarization internal resistance based on the basic input information characteristics to calculate real-time heat generation power, combine the convective heat transfer coefficient to calculate the equipment heat dissipation power, and calculate the net heat accumulation rate based on the real-time heat generation power and the equipment heat dissipation power. S2: Calculate the ratio of the net heat accumulation rate to the set equivalent heat capacity parameter to obtain the predicted temperature rise rate, evaluate the highest single-unit temperature and the current state of charge to obtain the temperature deviation state and the power deviation state, and classify each of the energy storage cabinets into the normal load state, critical load state, limited recovery state or prohibited heavy load state according to the predicted temperature rise rate, the temperature deviation state and the power deviation state, and generate the operation state classification result. S3: Filter the first type of energy storage cabinets that belong to the critical load-bearing state according to the operation state division results. Generate the maximum allowable heat generation power based on the preset maximum allowable temperature rise rate of the first type of energy storage cabinet and the heat dissipation power of the equipment. Calculate the upper limit of the evaluated output power. Filter the second type of energy storage cabinets that belong to the limited recovery state according to the operation state division results. Extract the actual maintenance power of the second type of energy storage cabinet as the locked upper limit of output power. Establish cabinet hierarchical constraint mapping association to obtain the upper limit of the load-bearing output power. S4: Calculate the absolute power margin by the difference between the upper limit of the load capacity and the current actual output capacity; The thermal state health factor is calculated based on the highest single-unit temperature and the current state of charge. The remaining safety margin matrix of the entire station is constructed based on the absolute power margin and the thermal state health factor. S5: Calculate the total power deficit and allocation ratio coefficient of the power station based on the remaining safety margin matrix of the entire station and the total power dispatching instruction of the power station, obtain the preliminary target incremental power, superimpose the preliminary target incremental power onto the current actual output to generate the expected power instruction, obtain the truncation limit instruction by selecting the smaller value between the expected power instruction and the upper limit of the load capacity, and adjust the power change rate of the truncation limit instruction using a preset slope limiter to obtain the target power execution instruction.

2. The intelligent management method for energy storage cabinets according to claim 1, characterized in that, The specific steps of S1 are as follows: S111: Collect basic operating status characteristics within the power station control cycle. The basic operating status characteristics include the highest single cell temperature of the battery, the current state of charge, the current charging and discharging current, the DC bus voltage, the speed and frequency of the cooling equipment, and the current actual output. Apply a sliding window filter to the basic operating status characteristics to calculate the truncated average value within the preset sampling window. Smooth the truncated average value according to the first-order low-pass filtering algorithm to construct the basic input information matrix. S112: Based on the relationship between the battery's historical operating temperature and internal resistance, a dynamic heat-sensitivity mapping matrix is ​​constructed to reflect the dynamic nonlinear mapping relationship between different states of charge and internal resistance. The current state of charge in the basic input information matrix is ​​called to match the internal resistance feature nodes in the dynamic heat-sensitivity mapping matrix to extract the dynamic polarization internal resistance. Based on the current charging and discharging current and the dynamic polarization internal resistance in the basic input information matrix, the product of the square parameter of the current charging and discharging current and the dynamic polarization internal resistance is calculated to obtain the real-time heat generation power. S113: Obtain the effective heat dissipation area of ​​the equipment and the set ambient temperature difference, map the energy storage cabinet as a heat accumulation model heating node, obtain the convective heat transfer coefficient of the heat accumulation model heating node according to the rotation speed frequency of the cooling equipment, calculate the product of the convective heat transfer coefficient, the effective heat dissipation area and the set ambient temperature difference to obtain the equipment heat dissipation power, and calculate the net heat accumulation rate according to the real-time heat generation power and the equipment heat dissipation power.

3. The intelligent management method for energy storage cabinets according to claim 2, characterized in that, The process of obtaining the convective heat transfer coefficient of the heating node in the heat accumulation model based on the rotational speed and frequency matching of the cooling equipment is as follows: By adjusting the cooling equipment to multiple fixed speed nodes under a preset standard test environment and collecting the surface heat flux data of the energy storage cabinet, a convective heat transfer mapping model is constructed through correlation calculation. Extract the rotational speed frequency of the cooling equipment and import the rotational speed frequency of the cooling equipment into the convection heat transfer mapping model; Interpolation is performed on the discrete data nodes within the convection heat transfer mapping model to extract the output heat transfer values ​​corresponding to the rotational speed frequency of the cooling equipment. The output heat transfer value is set as the convective heat transfer coefficient of the heating node in the heat accumulation model; The specific process for calculating the net heat accumulation rate is as follows: Calculate the power difference between the real-time heat generation power and the heat dissipation power of the equipment; The thermal conductivity of the internal support structure material of the energy storage cabinet is obtained and a preset reference thermal conductivity is obtained. A heat accumulation correction coefficient is set based on the ratio of the thermal conductivity of the internal support structure material of the energy storage cabinet to the preset reference thermal conductivity. Calculate the product of the power value difference and the heat accumulation correction factor, and use the product as the net heat accumulation rate.

4. The intelligent management method for energy storage cabinets according to claim 1, characterized in that, The specific steps of S2 are as follows: S211: Obtain the set equivalent heat capacity parameter of the battery module, calculate the ratio of the net heat accumulation rate to the set equivalent heat capacity parameter, and obtain the predicted temperature rise rate. S212: Call the basic input information matrix, extract the highest single-cell temperature and the current state of charge, obtain the preset temperature warning line, the preset forced load reduction line and the preset state of charge limit, compare the numerical deviation of the highest single-cell temperature with the preset temperature warning line and the preset forced load reduction line to obtain the temperature deviation state, and compare the numerical deviation of the current state of charge with the preset state of charge limit to obtain the charge deviation state. S213: Combine the predicted temperature rise rate, the temperature deviation state, and the power deviation state to determine the thermodynamic constraint boundary state of the energy storage cabinet. Based on the thermodynamic constraint boundary state, classify each of the energy storage cabinets into normal load-bearing state, critical load-bearing state, limited recovery state, or prohibited heavy load state, establish an operational quadrant distribution mapping association, and generate the operational state classification result of each of the energy storage cabinets.

5. The intelligent management method for energy storage cabinets according to claim 2, characterized in that, The specific steps of S3 are as follows: S311: Filter the first type of energy storage cabinet that belongs to the critical load state according to the operation state division result, obtain the preset maximum allowable temperature rise rate matched by the first type of energy storage cabinet, call the set equivalent heat capacity parameter and the equipment heat dissipation power, calculate the product of the set equivalent heat capacity parameter and the preset maximum allowable temperature rise rate, and add the equipment heat dissipation power to the product to generate the maximum allowable heat generation power. S312: Calculate the square root of the ratio between the maximum allowable heat generation power and the dynamic polarization internal resistance to obtain the maximum allowable discharge current corresponding to the first type of energy storage cabinet. Call the basic input information matrix to extract the DC-side bus voltage. Calculate the product of the maximum allowable discharge current and the DC-side bus voltage to obtain the upper limit of the evaluated output. S313: Filter the second type of energy storage cabinets that belong to the restricted recovery state according to the operation state division result, read the restricted constraint data packets reported by the battery management device corresponding to the second type of energy storage cabinet, extract the actual maintenance power recorded in the restricted constraint data packets, set the actual maintenance power as the locked output limit corresponding to the second type of energy storage cabinet, aggregate the evaluated output limit and the locked output limit corresponding to different quadrants, establish cabinet level constraint mapping association, and obtain the upper limit of the output that can be accepted.

6. The intelligent management method for energy storage cabinets according to claim 5, characterized in that, The process of obtaining the maximum allowable discharge current corresponding to the first type of energy storage cabinet is as follows: Obtain the preset limit discharge current and stable operating current, and set the discharge safety correction coefficient based on the ratio of the limit discharge current to the stable operating current. Calculate the numerical ratio of the maximum allowable heat generation power to the dynamic polarization internal resistance; Perform a square root operation on the numerical ratio to extract the root value data; Calculate the product of the root value data and the discharge safety correction coefficient to determine the maximum allowable discharge current corresponding to the first type of energy storage cabinet; The process of obtaining the upper limit of the load capacity is as follows: Construct a mapping form that includes a hardware identification field and an output limit field; Extract the hardware code of the first type of energy storage cabinet and fill it into the hardware identifier field, and store the evaluated output limit into the corresponding output limit field; Extract the hardware code of the second type of energy storage cabinet and fill it into the hardware identifier field, and store the locked output limit into the corresponding output limit field; Extract the record values ​​of all the output limit fields in the mapping form to obtain the upper limit of the capacity to be accepted.

7. The intelligent management method for energy storage cabinets according to claim 1, characterized in that, The specific steps of S4 are as follows: S411: Call the basic input information matrix, extract the current actual output, calculate the difference between the upper limit of the load-bearing output and the current actual output, and obtain the absolute power margin; S412: Obtain a preset set limit temperature and a set temperature range; call the basic input information matrix to extract the highest single-unit temperature; calculate the temperature value deviation based on the set limit temperature and the highest single-unit temperature; calculate the ratio of the temperature value deviation to the set temperature range to obtain the temperature distance ratio; call the basic input information matrix to extract the current state of charge; obtain a preset state of charge limit; calculate the energy value deviation based on the current state of charge and the preset state of charge limit to obtain a set energy range; calculate the ratio of the energy value deviation to the set energy range to obtain the state of charge distance ratio; obtain a preset temperature margin weight and a preset state of charge margin weight; calculate a first product based on the preset temperature margin weight and the temperature distance ratio; calculate a second product based on the preset state of charge margin weight and the state of charge distance ratio; calculate the sum of the first product and the second product to generate a thermal state health factor. S413: Calculate the product of the absolute power margin and the thermal health factor to obtain the comprehensive power margin, obtain the cabinet identifier, aggregate the operating status division result, the upper limit of the capacity, the current actual output, the absolute power margin, the thermal health factor and the comprehensive power margin based on the cabinet identifier, establish a hierarchical data mapping structure, and generate the total station remaining safety margin matrix.

8. The intelligent management method for energy storage cabinets according to claim 7, characterized in that, The specific steps of S5 are as follows: S511: Obtain the total power dispatch instruction of the power station, call the operating status division result in the remaining safety margin matrix of the entire station, filter the available energy storage cabinets in the normal load-bearing state and the critical load-bearing state, extract the upper limit of the capacity to be carried and the comprehensive power margin corresponding to the available energy storage cabinets, calculate the sum for all the upper limits of the capacity to be carried to obtain the total safe output of the entire station, calculate the numerical difference between the total power dispatch instruction of the power station and the total safe output of the entire station to obtain the total power deficit of the power station, and calculate the sum for all the comprehensive power margins to obtain the total comprehensive margin of the entire station; S512: Calculate the ratio of the comprehensive power margin of a single available energy storage cabinet to the total comprehensive power margin of the entire station to obtain the allocation ratio coefficient; calculate the product of the total power deficit of the power station and the allocation ratio coefficient to obtain the preliminary target incremental power; extract the current actual output corresponding to the available energy storage cabinet; calculate the sum of the current actual output and the preliminary target incremental power to obtain the expected power command; select the minimum value between the expected power command and the upper limit of the acceptable output to obtain the truncation limit command; S513: Extract the thermal state health factor corresponding to the available energy storage cabinet, obtain a preset slope limiter, match the maximum power change rate within the preset slope limiter according to the thermal state health factor, and adjust the cutoff limit command according to the maximum power change rate using the preset slope limiter to obtain the target power execution command.

9. The intelligent management method for energy storage cabinets according to claim 8, characterized in that, The process of obtaining the truncation restriction instruction is as follows: Compare the expected power command with the numerical value of the maximum acceptable output power; When it is determined that the expected power command is greater than the maximum acceptable output power, the maximum acceptable output power is assigned to the cut-off limit command; When it is determined that the expected power command is not greater than the maximum output capacity, the expected power command is assigned to the cut-off limit command; The process of adjusting the cutoff limit command according to the maximum power change rate using the preset slope limiter is specifically as follows: Obtain the historical output power of the available energy storage cabinet in the previous control cycle; The absolute value of the difference between the cutoff limit command and the historical output power is calculated as the power command change. Obtain the hardware response delay time of the energy storage converter and the set standard control cycle, and map and set the command buffer coefficient according to the ratio of the hardware response delay time of the energy storage converter to the set standard control cycle. The dynamic limiting gradient is obtained by multiplying the maximum power change rate by the command buffer coefficient. Compare the magnitude of the power command change with the value of the dynamic limiting gradient; When it is determined that the change in the power command is greater than the dynamic limiting gradient, the sum of the historical output power and the dynamic limiting gradient is calculated, and the sum is set as the target power execution command. When it is determined that the change in the power command is not greater than the dynamic limiting gradient, the truncation limiting command is used as the target power execution command.

10. An intelligent management system for energy storage cabinets, characterized in that, The system is used to implement the intelligent management method for energy storage cabinets as described in any one of claims 1-9, the system comprising: The multi-dimensional feature analysis module collects and processes the basic operating status features within the power plant control cycle, constructs a basic input information matrix, extracts dynamic polarization internal resistance based on the basic input information features to calculate real-time heat generation power, calculates equipment heat dissipation power in combination with convective heat transfer coefficient, and calculates net heat accumulation rate based on the real-time heat generation power and the equipment heat dissipation power. The load condition assessment module calculates the ratio of the net heat accumulation rate to the set equivalent heat capacity parameter to obtain the predicted temperature rise rate, assesses the highest single-unit temperature and the current state of charge to obtain the temperature deviation state and the power deviation state, and classifies each of the energy storage cabinets into normal load state, critical load state, limited recovery state or prohibited heavy load state according to the predicted temperature rise rate, the temperature deviation state and the power deviation state, and generates the operation state classification result. The capacity boundary estimation module filters out the first type of energy storage cabinets that belong to the critical load-bearing state based on the operation state division results. It generates the maximum allowable heat generation power based on the preset maximum allowable temperature rise rate of the first type of energy storage cabinet and the heat dissipation power of the equipment, and calculates the upper limit of the estimated output. It then filters out the second type of energy storage cabinets that belong to the limited recovery state based on the operation state division results, extracts the actual maintenance power of the second type of energy storage cabinet as the locked upper limit of output, establishes cabinet hierarchical constraint mapping association, and obtains the upper limit of the load-bearing capacity. The whole-site resource aggregation module calculates the difference between the upper limit of the capacity to be carried and the current actual output to obtain the absolute power margin; calculates the thermal state health factor based on the highest unit temperature and the current state of charge; and constructs the whole-site remaining safety margin matrix based on the absolute power margin and the thermal state health factor. The flexible collaborative allocation module calculates the total power deficit and allocation ratio coefficient of the power station based on the remaining safety margin matrix of the entire station and the total power dispatching command of the power station, obtains the preliminary target incremental power, superimposes the preliminary target incremental power onto the current actual output to generate the expected power command, obtains the truncation limit command by selecting the smaller value between the expected power command and the upper limit of the load capacity, and adjusts the power change rate of the truncation limit command using a preset slope limiter to obtain the target power execution command.

Images