Self-adaptive liquid cooling heat dissipation device and method for land energy storage container

Abstract

The invention discloses a land energy storage container self-adaptive liquid cooling heat dissipation device and method, the device comprises a liquid cooling unit, a self-adaptive circulation pipeline, a variable flow pump valve group, a distributed temperature sensing module and a BMS main control module, through data monitoring, model establishment, thermal load prediction, selection of a proper cooling mode, and introduction of reasonable compensation, the self-adaptive liquid cooling heat dissipation of the land energy storage container is realized. And the variable-frequency circulating pump and the valve body are adjusted and combined for execution. Through multi-parameter fusion control, dynamic impedance matching and intelligent algorithm optimization, a cooperative control mode of a variable orifice regulating valve and a variable frequency circulating pump is innovatively adopted, and dynamic matching of cooling loop impedance is achieved; through temperature field modeling, heat dissipation demand grading prediction, a cooling parameter self-optimization algorithm and fault self-diagnosis, the balance degree of the temperature field is effectively improved, heat dissipation energy consumption is reduced, environmental adaptability is enhanced, and meanwhile the method has the advantages of being high in reliability, wide in environmental adaptability and low in operation and maintenance cost.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of energy storage container heat dissipation, more particularly to a land energy storage container adaptive liquid cooling heat dissipation device and method. BACKGROUND

[0002] With the rapid development of renewable energy, energy storage boxes are increasingly widely used in the fields of power supply and electric transportation. However, a large amount of heat is generated during the charging and discharging of the battery pack. If the heat cannot be dissipated in time and effectively, it will lead to performance degradation, shortened service life, and even safety accidents.

[0003] Traditional land energy storage box heat dissipation methods usually use air cooling and liquid cooling. The existing land energy storage container liquid cooling system usually uses fixed flow or partition control mode.

[0004] This heat dissipation method has the following problems: 1. The cooling liquid flow distribution does not match the heat load of the battery cluster, resulting in local overheating or overcooling; 2. The traditional PID control response is lagging, which cannot adapt to dynamic charging and discharging conditions; 3. The imbalance of pipe impedance causes waste of pump energy consumption, and the system reliability is low. SUMMARY

[0005] In view of the deficiencies in the prior art, the purpose of the present application is to provide a land energy storage container adaptive liquid cooling heat dissipation device and method to solve one or more of the above problems.

[0006] To achieve the above purpose, the present application provides the following technical solutions:

[0007] A land energy storage container adaptive liquid cooling heat dissipation device, comprising

[0008] A liquid cooling unit is arranged outside the battery cabin of the box body between the liquid inlet pipeline and the liquid outlet pipeline;

[0009] An adaptive circulation pipeline, including a first-stage liquid cooling pipeline, a second-stage liquid cooling pipeline and a third-stage liquid cooling pipeline, is connected to the liquid inlet pipeline and the liquid outlet pipeline, respectively;

[0010] A variable flow pump valve group is arranged in the adaptive circulation pipeline;

[0011] A battery cluster is evenly arranged in the battery cabin, and each battery pack is provided with a micro-channel liquid cooling plate at the bottom, and the micro-channel liquid cooling plate is connected to the third-stage liquid cooling pipeline;

[0012] A distributed temperature sensing module is arranged on the surface of each battery cluster in the battery cabin and at the key nodes of the adaptive circulation pipeline;

[0013] A BMS main control module, including a signal sending and receiving unit, a temperature field modeling unit and a dynamic control unit,

[0014] The signal transmitting unit is electrically connected to the temperature field modeling unit, the fuzzy PID algorithm unit and the fault diagnosis unit respectively, the temperature field modeling unit is electrically connected to the fuzzy PID algorithm unit and the fault diagnosis unit respectively,

[0015] The signal transmitting unit receives the distributed temperature sensing module data signal and sends an execution signal to the liquid cooling unit and the variable flow valve pump group;

[0016] The temperature field modeling unit models and predicts through the data collected by the distributed temperature sensing module, reasonably switches between low load, medium load and high load modes, and issues adjustment instructions to the liquid cooling unit and / or the variable flow valve pump group, activates the fuzzy PID algorithm unit in the medium load mode, and activates the fault diagnosis unit in the abnormal operation;

[0017] The fuzzy PID algorithm unit is activated in the medium load mode, iteratively optimizes the PID parameters and issues adjustment instructions to the variable flow valve pump group;

[0018] The fault diagnosis unit analyzes the running condition and triggers the redundant protection mechanism and / or auxiliary cooling mechanism as needed and alarms when abnormal.

[0019] Further, the first liquid cooling pipeline, the second liquid cooling pipeline and the third liquid cooling pipeline are bidirectional pipelines provided with liquid inlets and liquid outlets;

[0020] The first liquid cooling pipeline is arranged at the bottom of the battery cabin of the box body and is connected to the liquid inlet pipeline and the liquid outlet pipeline respectively;

[0021] The second liquid cooling pipeline is arranged between the battery clusters in the battery cabin and is connected to the first liquid cooling pipeline;

[0022] The third liquid cooling pipeline is arranged between the battery clusters in the battery cabin and is connected to the stable channel liquid cooling plate at the bottom of the battery cluster and the second liquid cooling pipeline respectively;

[0023] All the liquid cooling pipelines adopt an asymmetric topology structure, the pipeline length is negatively correlated with the thermal load of the corresponding battery cluster, and a pressure sensor is arranged at the end of the pipeline;

[0024] The liquid inlet pipeline is provided with a total liquid inlet pipe on the side of the box body, the liquid outlet pipeline is provided with a total liquid outlet pipe on the side of the box body, and the liquid inlet pipeline, the liquid outlet pipeline, the first liquid cooling pipeline, the second liquid cooling pipeline and the third liquid cooling pipeline are all corrosion-resistant DN50 stainless steel pipes.

[0025] Further, the distributed temperature sensing module includes a thin film thermocouple arranged on the surface of the battery cluster, a temperature sensor arranged at the inlet and outlet of the battery pack micro-channel cooling plate, and a temperature and humidity sensor at a conventional temperature and humidity detection position in the battery cabin.

[0026] Furthermore, the variable flow pump valve assembly includes a variable orifice regulating valve and a pressure balancing valve located on the secondary liquid cooling pipeline. The variable orifice regulating valve consists of a gear pump driven by a stepper motor and a proportional regulating valve. The orifice diameter of the variable orifice regulating valve can be adjusted from 0.5 to 5 mm.

[0027] Liquid cooling units include compressors, condensers, evaporators, and variable frequency circulating pumps.

[0028] The heat dissipation method based on the above-mentioned adaptive liquid cooling heat dissipation device for land-based energy storage containers includes the following steps:

[0029] Sa, data acquisition, obtains surface temperature distribution data and environmental parameters of battery clusters through a distributed temperature sensing module;

[0030] Sb: Based on the data collected in step Sa, a three-dimensional temperature field model is established, and an LSTM neural network is applied to predict the trend of heat load change.

[0031] Sc. Based on the model established in step Sb and the predicted trend, select a suitable cooling mode to complete the dynamic control decision.

[0032] Sd, Execute the dynamic control decision completed in step Sc to the variable frequency circulating pump, variable flow valve pump group and liquid cooling unit;

[0033] SE monitors the entire battery compartment, performs real-time fault diagnosis, and provides timely alarms and records.

[0034] Furthermore, step Sa is further divided into

[0035] Sa1, collects surface temperature distribution data T of the battery cluster through a distributed temperature sensing module using a thin-film thermocouple array. i,j This forms a surface temperature matrix for the battery clusters;

[0036] Sa2, The inlet temperature T of the battery pack microchannel liquid cooling plate is obtained through the temperature sensor of the distributed temperature sensing module. in and outlet temperature T out Calculate the temperature difference;

[0037] Sa3, The ambient temperature T inside the battery compartment is collected by the temperature and humidity sensor of the distributed temperature sensing module. env and moderate H;

[0038] In step Sa1, T i,j In this context, i represents the battery cluster number, j represents the location of a single battery cell, and the data acquisition frequency is 1Hz.

[0039] In step Sa2, the temperature difference ΔT = T out -T in .

[0040] Further, step Sb is further divided into

[0041] Sb1, dividing the battery cluster three-dimensional grid based on the finite element analysis method, combining the heat conduction equation to calculate the temperature field distribution;

[0042] Sb2, inputting the existing charging and discharging current data of the system, the collected battery cabin environmental temperature and humidity data, the temperature difference data of the battery pack micro-channel cooling plate, and the battery cluster surface temperature matrix data to the long short-term memory LSTM neural network to form historical data;

[0043] According to the historical data, the thermal load change trend in the next 5-10 minutes is predicted;

[0044] Sb3, outputting the thermal load level and the thermal load growth rate,

[0045] Low load level, temperature standard deviation range ΔT < 2℃, thermal load growth rate r < 5% / min;

[0046] Medium load level, temperature standard deviation range 2℃≤ΔT < 4℃, thermal load growth rate 5% / min≤r < 15% / min;

[0047] High load level, temperature standard deviation range ΔT≥4℃, thermal load growth rate r≥15% / min.

[0048] Further, step Sc is further divided into

[0049] Sc1.1, low load mode: reduce the variable frequency circulating pump speed to 800-1500 rpm, reduce the total flow of the cooling liquid;

[0050] Close the variable orifice regulating valve to an aperture < 1mm;

[0051] Sc1.2, medium load mode: activate the fuzzy PID algorithm unit, dynamically adjust K p , K i , K d according to the temperature deviation e(k) and the temperature change rate Δe(k);

[0052] Call the genetic algorithm to optimize the flow distribution of the liquid cooling pipe, generate 10 groups of candidate schemes, and iteratively optimize the minimum energy consumption and minimum temperature difference;

[0053] Sc1.3, high load mode: make the compressor of the liquid cooling unit run at full power, set the cooling liquid temperature to 15℃, and open all the variable orifice regulating valves of the liquid cooling pipeline to an aperture ≥4mm to maximize the cooling liquid flow;

[0054] In step Sc1.2, according to the corresponding branch flow Q i , valve opening D and valve inlet and outlet pressure difference ΔPi According to the formula, the target flow rate of pipeline impedance matching is calculated

[0055]

[0056] Where, ρ is the density of the coolant, constant 3.6x10 3 is the dimension conversion coefficient;

[0057] In order to adapt to the change of pipeline impedance, pressure compensation is introduced l

[0058] Q i,actual = Q i + K bal · (P main -P ref )

[0059] Where, K bal is the pressure balance gain, P main is the real-time pressure of the main circuit, and P ref is the target pressure obtained by genetic algorithm;

[0060] The input variables of fuzzy PID algorithm are the temperature difference e(k) between the target temperature and the actual temperature of the battery surface, the difference Δe(k) between the current time and the last time, and the environmental humidity f(H) as a correction factor;

[0061] The fuzzy set of the membership function e(k) of the fuzzy process:

[0062] {Negative Big (NB), Negative Small (NS), Zero (ZO), Positive Small (PS), Positive Big (PB)}, triangular membership function is adopted;

[0063] The fuzzy set of Δe(k):

[0064] {Fast Down (FD), Slow Down (SD), Stable (ST), Slow Up (SU), Fast Up (FU)}, Gaussian membership function is adopted;

[0065] Finally, the fuzzy output is converted into accurate PID parameters by using the COG method:

[0066]

[0067] Where, u i is the activation degree of each rule,

[0068] The genetic algorithm randomly generates 10 sets of candidate solutions of flow distribution scheme, each set containing the proportion of each branch flow, and the constraint condition is:

[0069] ΣQ i = Q total

[0070] The fitness function is:

[0071]

[0072] Where P pump is the actual power of the circulating pump, the safety threshold ΔT threshold = 4℃,

[0073] Select the individual that retains 30% of the fitness, when the fitness changes <1% for 3 consecutive generations, or the maximum number of iterations is 50 generations, terminate the iteration;

[0074] In step Sc1.3, when the coolant temperature is lower than the ambient temperature, turn on the electric heating to prevent freezing

[0075] Further, step Sd is further divided into

[0076] Sd1, the variable frequency circulating pump adjusts the speed according to the target total flow Q total , introduces fuzzy PID output correction;

[0077] Sd2, the variable throttling hole regulating valve controls the valve core position through the stepper motor;

[0078] Sd3, the liquid cooling unit adjusts the compressor power and the expansion valve opening to maintain the coolant temperature;

[0079] The adjustment accuracy of the variable frequency circulating pump speed in step Sd1 is ±10 rpm;

[0080] The determination of the variable frequency circulating pump speed is determined according to the heat load demand, pipeline impedance and energy efficiency constraints, and the core calculation formula is

[0081]

[0082] Where Q total is the total flow demand of the coolant per minute, the maximum rated flow of the pump per minute Q max determined by the heat load prediction model, N max is the rated maximum speed of the pump; k1, k2 are weight coefficients, k1 = 0.7, k2 = 0.3; ΔT tactual is the actual coolant inlet and outlet temperature difference, ΔT target is the target temperature difference;

[0083] The fuzzy PID output correction can improve the response speed and energy efficiency, and the correction formula is

[0084] N final = N + ΔN PID

[0085]

[0086] wherein e N = N target -N actual , K p , K i , K d is the PID parameter generated by the fuzzy rule;

[0087] The resolution of the opening degree of the variable orifice regulating valve in step Sd2 is 0.1 mm;

[0088] The opening degree D of the variable orifice regulating valve is dynamically calculated through liquid cooling pipeline flow distribution and impedance matching, and the calculation formula is

[0089]

[0090] wherein Di is the orifice diameter corresponding to the branch valve opening degree, Q i is the target flow per minute of the corresponding branch, ΔP i is the pressure drop of the corresponding branch, which is obtained in real time through a sensor; C v is the flow coefficient, which is determined by the valve body structure;

[0091] To solve the coupling interference of multiple branches, pressure balance compensation is introduced

[0092] D i,final = D i + ΔD i

[0093] ΔD i = K bal · (P main -P i )

[0094] wherein P main is the main circuit pressure, P i is the inlet pressure of the corresponding branch, and K bal is the balance gain coefficient;

[0095] To avoid the liquid cooling liquid from shaking in the pipeline, the control instructions of the variable frequency pump and the regulating valve need to be solved synchronously. When the load is high, the circulating pump speed is adjusted first, and then the valve opening degree is optimized. When the load is low, the valve throttling is closed first, and then the pump speed is reduced. The dynamic decoupling algorithm of the instruction is:

[0096]

[0097] wherein α and β are decoupling coefficients, which are updated in real time through pressure feedback;

[0098] In step Sd3, the liquid cooling unit maintains the cooling liquid temperature at the set value ±5℃.

[0099] Further, step Se is also divided into

[0100] Se1.1, flow anomaly detection: actual flow Q of liquid cooling pipeline i Deviation from the set value for a certain period of time, determine the blockage or leakage, close the corresponding valve, switch to standby redundant pipeline;

[0101] Se1.2, temperature consistency detection: battery cluster temperature standard deviation ΔT abnormal for a certain period of time, determine the sensor failure or local overheating, start auxiliary cooling scheme, and mark a sensor;

[0102] Se2, alarm and record,

[0103] Se2.1, first level warning, yellow LED light, slight abnormality, system automatic repair;

[0104] Se2.2, second level warning, red LED light and buzzer, serious fault, need manual intervention;

[0105] Se2.3, record fault time, type and processing log to the memory of the main controller module;

[0106] Actual flow Q in step Se1.1 i Deviation from the set value ± 15% and lasts more than 30s, determine the blockage or leakage;

[0107] Battery cluster temperature standard deviation ΔT in step Se1.2 is greater than or equal to 3℃ and lasts more than 1min, determine the sensor failure or local overheating;

[0108] The auxiliary cooling scheme is to start the forced air cooling auxiliary cooling of the adjacent area.

[0109] In summary, the present application has the following beneficial effects: through multi-parameter fusion control, dynamic impedance matching and intelligent algorithm optimization, the variable orifice regulating valve and the variable frequency circulating pump are innovatively used in the collaborative control mode, the dynamic matching of the cooling loop impedance is realized; through temperature field modeling, heat dissipation demand hierarchical prediction, cooling parameter self-optimization algorithm and fault self-diagnosis, the temperature field uniformity is effectively improved, the heat dissipation energy consumption is reduced, and the environmental adaptability is enhanced, the problems of traditional liquid cooling system response lag, poor uniformity and high energy consumption are solved, and the present application has the characteristics of high reliability, wide environmental adaptability and low operation and maintenance cost, and is especially suitable for the thermal management needs of high power density energy storage containers, and provides an efficient thermal management solution for high power density energy storage containers. BRIEF DESCRIPTION OF DRAWINGS

[0110] Fig. 1 The main structure schematic diagram of one embodiment provided by the present application;

[0111] Fig. 2This is a schematic diagram showing the positions of the main outlet pipe and the main inlet pipe in one embodiment of the present invention;

[0112] Fig. 3 A system flowchart of one embodiment of the present invention;

[0113] Fig. 4 This is a control relationship diagram for one embodiment of the present invention.

[0114] In the diagram: 1. Liquid cooling unit; 2. Battery pack; 3. Distributed temperature sensing module; 4. Three-stage liquid cooling pipeline; 5. Variable flow pump valve assembly; 6. Secondary liquid cooling pipeline; 7. Primary liquid cooling pipeline; 8. Main liquid outlet pipe; 9. Main liquid inlet pipe. Detailed Implementation

[0115] Example:

[0116] The following is in conjunction with the appendix Figs. 1-4 The present invention will be described in further detail below.

[0117] An adaptive liquid cooling heat dissipation device for land-based energy storage containers, such as Figs. 1-2 As shown, including

[0118] Liquid cooling unit 1 is located outside the battery compartment of the enclosure, between the inlet and outlet liquid lines. Liquid cooling unit 1 includes a compressor, condenser, evaporator, and variable frequency circulating pump. Preferably, it adopts an integrated variable frequency circulating pump, a double-compressor refrigeration unit, and a plate heat exchanger. The refrigerant is R410A, and the coolant is a 50% ethylene glycol solution. Liquid cooling unit 1 can provide coolant and regulate temperature. Liquid cooling unit 1 effectively dissipates the heat generated during the charging and discharging of battery pack 2 through circulating coolant, ensuring that battery pack 2 is within a suitable operating temperature range and avoiding shortened battery pack 2 life or performance degradation.

[0119] Adaptive circulation pipeline, including parallel first-stage liquid cooling pipeline 7, second-stage liquid cooling pipeline 6 and third-stage liquid cooling pipeline 4, all corresponding to the liquid inlet pipeline and liquid outlet pipeline respectively; The first-stage liquid cooling pipeline 7, the second-stage liquid cooling pipeline 6 and the third-stage liquid cooling pipeline 4 are all bidirectional pipelines provided with liquid inlet and liquid outlet; The first-stage liquid cooling pipeline 7 is arranged at the bottom of the battery cabin of the box body, and is connected to the liquid inlet pipeline and the liquid outlet pipeline respectively; The second-stage liquid cooling pipeline 6 is arranged between the battery clusters in the battery cabin, and is connected to the first-stage liquid cooling pipeline 7; The third-stage liquid cooling pipeline 4 is arranged between the battery clusters in the battery cabin, and is connected to the micro-channel liquid cooling plate at the bottom of the battery cluster and the second-stage liquid cooling pipeline 6 respectively; All the liquid cooling pipelines adopt an asymmetric topological structure, that is, a serpentine loop cooling pipeline with opposite loop directions, the water inlet of the upper pipeline and the water outlet of the lower pipeline are arranged on the same side, the water outlet of the upper pipeline and the water inlet of the lower pipeline are arranged on the same side, and the upper and lower pipelines are composed of at least one hollow semicircular tube, the inner arc tube diameter is the same as the radius of the monomer battery pack 2, the pipeline length is negatively correlated with the heat load of the corresponding battery cluster, and a pressure sensor is arranged at the tail end; The liquid inlet pipeline is provided with a total liquid inlet pipe 9 on the side of the box body, the liquid outlet pipeline is provided with a total liquid outlet pipe 8 on the side of the box body, the liquid inlet pipeline, the liquid outlet pipeline, the first-stage liquid cooling pipeline 7, the second-stage liquid cooling pipeline 6 and the third-stage liquid cooling pipeline 4 are all DN50 stainless steel pipes with corrosion resistance; The cooling liquid is output from the liquid cooling unit 1, enters the main structure of the battery cabin through the total liquid inlet pipe 9 and the total liquid outlet pipe 8, is sequentially distributed to the second-stage liquid cooling pipeline 6 and the third-stage liquid cooling pipeline 4 through the first-stage liquid cooling pipeline 7, and is finally transported to the micro-channel liquid cooling plate at the bottom of the battery pack 2.

[0120] Variable flow pump valve group 5 arranged in the adaptive circulation pipeline; The variable flow pump valve group 5 includes a variable orifice adjusting valve and a pressure balance valve arranged on the second-stage liquid cooling pipeline 6, the variable orifice adjusting valve is composed of a gear pump driven by a stepping motor and a proportional adjusting valve, and can realize linear regulation of flow in response to a control signal; The variable orifice adjusting valve has an orifice diameter adjustment range of 0.5-5mm, a response time of ≤200ms, and a valve body material of corrosion-resistant aluminum alloy; Local heat area targeted heat dissipation is realized by dynamically adjusting the flow distribution of the cooling liquid.

[0121] The battery clusters are evenly arranged in the battery cabin, and each battery pack 2 is provided with a micro-channel liquid cooling plate at the bottom, and the micro-channel liquid cooling plate is connected to the third-stage liquid cooling pipeline 4; The micro-channel liquid cooling plate of the battery pack 2 adopts a composite structure of serpentine flow channel and tree-shaped flow channel, and independent electromagnetic valves are arranged between the sub-zones.

[0122] Distributed temperature sensing module 3 is arranged on the surface of each battery cluster in the battery cabin and the key nodes of the adaptive circulation pipeline, which is used to collect the temperature distribution of the battery cluster and the environmental parameters in real time; the distributed temperature sensing module 3 includes a thin film thermocouple arranged on the surface of the battery cluster, a temperature sensor arranged at the inlet and outlet of the micro-channel cooling plate of the battery pack 2, and a temperature and humidity sensor arranged at a conventional temperature and humidity detection position in the battery cabin; in addition, it also includes an infrared sensor, a fiber sensor lamp, and is installed at positions such as the ventilation opening of the container shell to monitor the external wind speed parameter. The distributed temperature sensor communicates with the BMS main control module through the communication wire harness, and the temperature sensor can collect the cell voltage and cell temperature of each battery in the battery pack 2 and transmit it back to the BMS main control module. The BMS main control module collects and monitors the cell temperature of each battery pack 2 temperature feedback, and issues instructions to each valve motor for opening and closing control.

[0123] The BMS main control module includes a signal receiving and sending unit, a temperature field modeling unit, and a dynamic control unit. The signal receiving and sending unit is electrically connected to the temperature field modeling unit, the fuzzy PID algorithm unit, and the fault diagnosis unit. The temperature field modeling unit is electrically connected to the fuzzy PID algorithm unit and the fault diagnosis unit. The signal receiving and sending unit receives data signals from the distributed temperature sensing module 3 and sends execution signals to the liquid cooling unit 1 and the variable flow valve pump group. The temperature field modeling unit models and predicts based on the data collected by the distributed temperature sensing module 3, reasonably switches between low load, medium load, and high load modes, and issues adjustment instructions to the liquid cooling unit 1 and / or the variable flow valve pump group. The fuzzy PID algorithm unit is activated in medium load mode, and the fault diagnosis unit is activated in abnormal operation. The fuzzy PID algorithm unit can input battery cluster surface temperature distribution data, environmental temperature and humidity, and historical thermal load data, establish dynamic weight coefficients based on temperature field gradients, calculate target heat dissipation power, and iteratively optimize PID parameters. The fuzzy PID algorithm unit sends speed instructions to the variable frequency circulating pump and opening degree instructions to the variable throttle hole regulating valve. The fault diagnosis unit analyzes the operation and triggers the redundant protection mechanism and / or auxiliary cooling mechanism as needed and alarms when abnormal. The fault diagnosis unit includes a flow anomaly detection module and a temperature consistency analysis module. The former detects pipeline blockage or leakage based on pressure-flow coupling relationship, and the latter identifies sensor failure or local overheating based on standard deviation threshold. The system automatically switches to redundant cooling mode and triggers an alarm signal. The system triggers a liquid leakage warning or pipeline blockage warning by detecting pump valve group current fluctuation, pipeline pressure, and cooling liquid conductivity changes. The system generates control instructions based on sensor data to cooperatively adjust the speed of the variable frequency circulating pump, the opening degree of the variable throttle hole, and the refrigeration power of the liquid cooling unit 1.

[0124] Based on the above-mentioned land energy storage container adaptive liquid cooling heat dissipation device matched heat dissipation method, mainly to control the constant temperature inside the box body, such as Figs. 3-4 As shown, comprising the following steps,

[0125] Sa, data acquisition,

[0126] Sa1, through the thin film thermocouple array of distributed temperature sensing module 3, the battery cluster surface temperature distribution data T i,j is collected,

[0127] In step Sa1, T i,j i represents the battery cluster number, j represents the monomer battery position, and the data acquisition frequency is 1Hz;

[0128] Sa2, through the temperature sensor of distributed temperature sensing module 3, the battery pack 2 microchannel liquid cooling plate inlet temperature T in And outlet temperature T out , calculate the temperature difference;

[0129] In step Sa2, the temperature difference ΔT = T out -T in .

[0130] Sa3, through the temperature and humidity sensor of distributed temperature sensing module 3, the environmental temperature T env And moderate H in the battery cabin are collected;

[0131] Sb, thermal field modeling and prediction,

[0132] Sb1, based on the finite element analysis method, the battery cluster three-dimensional grid is divided, and the temperature field distribution is calculated combined with the heat conduction equation;

[0133] In step Sb1, the battery cluster three-dimensional grid is divided according to 5mm x 5mm x 5mm;

[0134] Sb2, to the long short-term memory LSTM neural network, constantly input the existing charge and discharge current data of the system, the collected battery cabin environmental temperature and humidity data, the temperature difference data of the battery pack 2 microchannel cooling plate, and the battery cluster surface temperature matrix data, form historical data;

[0135] According to the historical data, the thermal load change trend in the future 5-10min is predicted;

[0136] In step Sb2, the input layer of LSTM neural network is the historical data with time window length T = 30min, the surface temperature matrix of battery cluster set and the microchannel cooling plate temperature difference are sampled every 5s, the charge and discharge current is sampled every 1Hz, and the environmental temperature and humidity are sampled every 10s;

[0137] The hyperparameters of the LSTM neural network are units = 256, batch size = 128, layer number = 2, epoch = 150, using RELU as the activation function, and Dropout = 2.

[0138] Sb3, output heat load level and heat load growth rate,

[0139] Low load level, temperature standard deviation range ΔT < 2℃, heat load growth rate r < 5% / min;

[0140] Medium load level, temperature standard deviation range 2℃ ≤ ΔT < 4℃, heat load growth rate 5% / min ≤ r < 15% / min;

[0141] High load level, temperature standard deviation range ΔT ≥ 4℃, heat load growth rate r ≥ 15% / min;

[0142] Sc, dynamic control decision,

[0143] Sc1, select control mode according to heat load level,

[0144] Sc1.1, low load mode: reduce the variable frequency circulating pump speed to 800-1500 rpm, reduce the total flow of cooling liquid;

[0145] Close the variable orifice regulating valve to an aperture <1mm;

[0146] Sc1.2, medium load mode: activate the fuzzy PID algorithm unit, dynamically adjust K p , K i , K d according to temperature deviation e(k) and temperature difference change rate Δe(k);

[0147] Call the genetic algorithm to optimize the flow distribution of the liquid cooling pipe, generate 10 groups of candidate schemes, and iteratively optimize with the goals of minimum energy consumption and minimum temperature difference;

[0148] In step Sc1.2, according to the corresponding branch flow Q i , valve opening D and valve inlet and outlet pressure difference ΔP i , the target flow of pipe impedance matching is calculated according to the formula

[0149]

[0150] Where ρ is the density of the cooling liquid, and the constant 3.6x103 is the dimension conversion coefficient;

[0151] In order to adapt to the change of pipe impedance, pressure compensation is introduced l

[0152] Q i,actual= Q i + K bal · (P main - P ref )

[0153] Where, K bal is the pressure balance gain, P main is the real-time pressure of the main circuit, P ref is the target pressure obtained by genetic algorithm;

[0154] The input variables of the fuzzy PID algorithm are the temperature difference e(k) between the target temperature and the actual temperature of the battery surface, the difference Δe(k) between the current time and the previous time, and the environmental humidity f(H) as a correction factor;

[0155] The fuzzy set of the membership function e(k) of the fuzzy process:

[0156] {negative big (NB), negative small (NS), zero (ZO), positive small (PS), positive big (PB)}, triangular membership function is adopted;

[0157] The fuzzy set of Δe(k):

[0158] {fast drop (FD), slow drop (SD), stable (ST), slow rise (SU), fast rise (FU)}, Gaussian membership function is adopted;

[0159] Finally, the fuzzy output is converted into accurate PID parameters by using the COG method:

[0160]

[0161] Where, u i is the activation degree of each rule,

[0162] The genetic algorithm randomly generates 10 sets of candidate solutions of flow distribution schemes through population initialization, each set containing the proportion of each branch flow, and the constraint conditions are:

[0163] ΣQ i = Q total

[0164] The fitness function is:

[0165]

[0166] Where, P pump is the actual power of the circulating pump, and the safety threshold ΔT threshold = 4℃,

[0167] The individuals with the top 30% fitness are selected, and the iteration is terminated when the fitness changes by less than 1% for three consecutive generations or the maximum iteration number of 50 generations is reached;

[0168] Sc1.3, high load mode: the compressor of the liquid cooling unit 1 is operated at full power, the cooling liquid set temperature is 15°C, all the variable orifice regulating valves of the liquid cooling pipeline are fully opened to an aperture ≥4mm, and the cooling liquid flow is maximized;

[0169] When the cooling liquid temperature is lower than the ambient temperature in step Sc1.3, the electric heating is turned on to prevent freezing.

[0170] Sd, execute control instructions,

[0171] Sd1, the variable frequency circulating pump adjusts the target total flow Q total Adjust the speed, introduce fuzzy PID output correction;

[0172] The adjustment accuracy of the variable frequency circulating pump speed in step Sd1 is ±10rpm;

[0173] The determination of the variable frequency circulating pump speed is determined according to the heat load demand, the pipeline impedance and the energy efficiency constraint, and the core calculation formula is

[0174]

[0175] Wherein, Q total is the total flow demand of the cooling liquid per minute, the rated maximum flow of the pump per minute Q max is determined by the heat load prediction model, N max is the rated highest speed of the pump; k1, k2 are weight coefficients, k1=0.7, k2=0.3; ΔT tactual is the actual cooling liquid inlet and outlet temperature difference, ΔT target is the target temperature difference;

[0176] The fuzzy PID output correction can improve the response speed and energy efficiency, and the correction formula is

[0177] N final =N+ΔN PID

[0178]

[0179] Wherein, e N =N target -N actual , K p , K i , K d are the PID parameters generated by the fuzzy rule;

[0180] Sd2, the variable orifice regulating valve controls the valve core position through the stepper motor;

[0181] The opening resolution of the variable orifice regulating valve in step Sd2 is 0.1mm;

[0182] The opening degree D of the variable orifice regulating valve is dynamically calculated through liquid cooling pipeline flow distribution and impedance matching, and the calculation formula is

[0183]

[0184] Wherein, Di is the orifice diameter corresponding to the branch valve opening, Q i is the target flow per minute of the corresponding branch, ΔP i is the pressure drop of the corresponding branch, which is obtained in real time through a sensor; C v is the flow coefficient, which is determined by the valve body structure;

[0185] To solve the coupling interference of multiple branches, pressure balance compensation is introduced

[0186] D i,final =D i +ΔD i

[0187] ΔD i =K bal ·(P main -P i )

[0188] Wherein, P main is the main circuit pressure, P i is the inlet pressure of the corresponding branch, and K bal is the balance gain coefficient;

[0189] In order to avoid the liquid cooling liquid from shaking in the pipeline, the control instructions of the variable frequency pump and the regulating valve need to be solved synchronously. When the load is high, the circulating pump speed is adjusted first, and then the valve opening is optimized. When the load is low, the valve throttling is closed first, and then the pump speed is reduced. The dynamic decoupling algorithm of the instruction is:

[0190]

[0191] Wherein, α and β are decoupling coefficients, which are updated in real time through pressure feedback;

[0192] Sd3, the liquid cooling unit 1 adjusts the compressor power and the expansion valve opening to maintain the cooling liquid temperature;

[0193] In step Sd3, the liquid cooling unit 1 maintains the cooling liquid temperature at the set value ±5℃.

[0194] Se, fault diagnosis and redundancy protection,

[0195] Se1, real-time monitoring and judgment

[0196] Se1.1, flow anomaly detection: the actual flow Q iDeviation from the set value for a certain period of time, determine the blockage or leakage, close the corresponding valve, switch to the standby redundant pipeline;

[0197] Actual flow Q in step Se1.1 i Deviation from the set value for a certain period of time, determine the blockage or leakage, close the corresponding valve, switch to the standby redundant pipeline;

[0198] Se1.2, temperature consistency detection: battery cluster temperature standard deviation ΔT abnormal for a certain period of time, determine the sensor failure or local overheating, start auxiliary cooling scheme, and mark a sensor;

[0199] Se1.2, temperature consistency detection: battery cluster temperature standard deviation ΔT abnormal for a certain period of time, determine the sensor failure or local overheating, start auxiliary cooling scheme, and mark a sensor;

[0200] The auxiliary cooling scheme is to start the forced air cooling auxiliary cooling of the adjacent area.

[0201] Se2, alarm and record,

[0202] Se2.1, first level warning, bright yellow LED light, slight abnormality, system automatic repair;

[0203] Se2.2, second level warning, bright red LED light and buzzer, serious fault, need manual intervention;

[0204] Se2.3, record fault time, type and processing log to the memory of the main controller module.

[0205] It should be noted that the specific embodiments are merely an explanation of the present application, and are not a limitation of the present application. Those skilled in the art can make non-creative modifications to the embodiments according to the needs after reading the specification, but as long as the modifications are within the scope of the claims of the present application, they are protected by the patent law.

Claims

1. An adaptive liquid cooling heat dissipation device for land energy storage container, characterized in that: The liquid cooling unit is arranged outside the battery cabin of the box body, between the liquid inlet pipeline and the liquid outlet pipeline; The adaptive circulation pipeline includes a first-stage liquid cooling pipeline, a second-stage liquid cooling pipeline and a third-stage liquid cooling pipeline, which are connected to the liquid inlet pipeline and the liquid outlet pipeline respectively; The variable flow pump valve group is arranged in the adaptive circulation pipeline; The battery clusters are arranged in the battery cabin, and each battery pack is provided with a micro-channel liquid cooling plate at the bottom, which is connected to the third-stage liquid cooling pipeline; The distributed temperature sensing module is arranged on the surface of each battery cluster in the battery cabin and the key nodes of the adaptive circulation pipeline; The BMS main control module includes a signal receiving and sending unit, a temperature field modeling unit and a dynamic control unit, The signal receiving and sending unit is electrically connected to the temperature field modeling unit, the fuzzy PID algorithm unit and the fault diagnosis unit, and the temperature field modeling unit is electrically connected to the fuzzy PID algorithm unit and the fault diagnosis unit, The signal receiving and sending unit receives the data signal of the distributed temperature sensing module and sends the execution signal to the liquid cooling unit and the variable flow valve pump group; The temperature field modeling unit models and predicts through the data collected by the distributed temperature sensing module, reasonably switches between low load, medium load and high load modes, and issues adjustment instructions to the liquid cooling unit and / or the variable flow valve pump group, activates the fuzzy PID algorithm unit in the medium load mode, and activates the fault diagnosis unit in the abnormal operation; The fuzzy PID algorithm unit is activated in the medium load mode, iteratively optimizes the PID parameters and issues adjustment instructions to the variable flow valve pump group; The fault diagnosis unit analyzes the running condition and triggers the redundancy protection mechanism and / or auxiliary heat dissipation mechanism as needed and alarms when the running condition is abnormal.

2. The adaptive liquid cooling heat sink for land-based energy storage containers of claim 1, wherein: The first-stage liquid cooling pipeline, the second-stage liquid cooling pipeline and the third-stage liquid cooling pipeline are bidirectional pipelines provided with liquid inlets and liquid outlets; The first-stage liquid cooling pipeline is arranged at the bottom of the battery cabin of the box body and connected to the liquid inlet pipeline and the liquid outlet pipeline; The second-stage liquid cooling pipeline is arranged between the battery clusters in the battery cabin and connected to the first-stage liquid cooling pipeline; The third-stage liquid cooling pipeline is arranged between the battery clusters in the battery cabin and connected to the micro-channel liquid cooling plate at the bottom of each battery cluster and the second-stage liquid cooling pipeline; All the liquid cooling pipelines adopt an asymmetric topology structure, the pipeline length is negatively correlated with the thermal load of the corresponding battery cluster, and a pressure sensor is arranged at the end of each pipeline; The liquid inlet pipeline is provided with a total liquid inlet pipeline on the side of the box body, the liquid outlet pipeline is provided with a total liquid outlet pipeline on the side of the box body, and the liquid inlet pipeline, the liquid outlet pipeline, the first-stage liquid cooling pipeline, the second-stage liquid cooling pipeline and the third-stage liquid cooling pipeline are all DN50 stainless steel pipelines resistant to corrosion.

3. The adaptive liquid cooling heat sink for land-based energy storage containers of claim 1, wherein: The distributed temperature sensing module includes a thin-film thermocouple arranged on the surface of the battery cluster, a temperature sensor arranged at the inlet and outlet of the micro-channel cooling plate of the battery pack, and a temperature and humidity sensor arranged at a conventional temperature and humidity detection position in the battery cabin.

4. The adaptive liquid cooling heat sink for energy storage container of claim 1, wherein: The variable flow pump valve group includes a variable orifice adjusting valve and a pressure balance valve arranged on the second-stage liquid cooling pipeline, the variable orifice adjusting valve is composed of a gear pump driven by a stepping motor and a proportional adjusting valve, and the variable orifice adjusting valve has an orifice diameter adjusting range of 0.5-5 mm; The liquid cooling unit includes a compressor, a condenser, an evaporator and a variable frequency circulating pump.

5. An adaptive liquid cooling method for a land energy storage container, characterized in that: The adaptive liquid cooling heat dissipation device of the land energy storage container according to any one of claims 1-4 comprises the following steps, Sa, data acquisition, obtains surface temperature distribution data and environmental parameters of battery clusters through a distributed temperature sensing module; Sb: Based on the data collected in step Sa, a three-dimensional temperature field model is established, and an LSTM neural network is applied to predict the trend of heat load change. Sc. Based on the model established in step Sb and the predicted trend, select a suitable cooling mode to complete the dynamic control decision. Sd, Execute the dynamic control decision completed in step Sc to the variable frequency circulating pump, variable flow valve pump group and liquid cooling unit; SE monitors the entire battery compartment, performs real-time fault diagnosis, and provides timely alarms and records.

6. The adaptive liquid cooling heat dissipation method for land energy storage container according to claim 5, characterized in that: Step Sa is further divided into Sa1, collecting battery cluster surface temperature distribution data T by a thin-film thermocouple array of a distributed temperature sensing module i,j , forming a battery cluster surface temperature matrix; Sa2. Obtain the battery pack micro-channel liquid cooling plate inlet temperature T through the temperature sensor of the distributed temperature sensing module in and outlet temperature T out , calculate the temperature difference; Sa3, the environmental temperature T in the battery cabin is collected by the temperature and humidity sensor of the distributed temperature sensing module env and moderate H; T in step Sa1 i,j i represents the battery cluster number, j represents the monomer battery position, and the data collection frequency is 1 Hz; The temperature difference ΔT = T out - T in .

7. The adaptive liquid cooling heat dissipation method for land energy storage container according to claim 5, characterized in that: Step Sb is further divided into Sb1. The three-dimensional mesh of the battery cluster is generated based on the finite element analysis method, and the temperature field distribution is calculated by combining the heat conduction equation. Sb2, continuously input the existing charging and discharging current data, collected battery compartment ambient temperature and humidity data, battery pack microchannel cooling plate temperature difference data, and battery cluster surface temperature matrix data into the Long Short-Term Memory (LSTM) neural network to form historical data; Based on historical data, predict the trend of heat load changes in the next 5 to 10 minutes; Sb3, output heat load level and heat load growth rate Low load level, temperature standard deviation range ΔT < 2℃, heat load growth rate r < 5% / min; Medium load level, temperature standard deviation range 2℃≤ΔT<4℃, heat load growth rate 5% / min≤r<15% / min; High load level, temperature standard deviation range ΔT≥4℃, heat load growth rate r≥15% / min.

8. The adaptive liquid cooling heat dissipation method for land energy storage container according to claim 5, characterized in that: Step Sc is further divided into Sc1.1, Low Load Mode: Reduce the speed of the variable frequency circulating pump to 800-1500 rpm to reduce the total coolant flow rate; Close the variable orifice regulating valve until the orifice diameter is <1mm; Sc1.2, medium load mode: activate fuzzy PID algorithm unit, dynamically adjust K according to temperature deviation e(k) and temperature change rate Δe(k) p , K i , K d ; A genetic algorithm is used to optimize the flow distribution of the liquid cooling pipeline, generating 10 candidate schemes, and iteratively optimizing them with the goals of minimizing energy consumption and minimizing temperature difference. Sc1.3, High Load Mode: The compressor of the liquid-cooled unit runs at full power, the coolant is set at 15℃, and all variable throttle orifice regulating valves of the liquid-cooled pipelines are opened to an orifice diameter ≥4mm to maximize the coolant flow. In step Sc1.2, the target flow rate Q i , the valve opening D, and the valve inlet and outlet pressure difference ΔP i The target flow rate of the pipe impedance matching is calculated according to the formula where p is the coolant density, the constant 3.6 x 10 3 is a dimension conversion factor; To accommodate changes in line impedance, pressure compensation is introduced l Q i,actual = Q i + K bal · (P main - P ref ) wherein K bal is the pressure balance gain, P main is the main circuit real-time pressure, P ref is the target pressure derived from the genetic algorithm; The input variables of the fuzzy PID algorithm are the temperature difference e(k) between the target temperature and the actual temperature of the battery surface, the difference Δe(k) between the current time and the previous time, and the ambient humidity f(H) as correction factors. The fuzzy set of the membership function e(k) of a fuzzy process: {Negative Large (NB), Negative Small (NS), Zero (ZO), Positive Small (PS), Positive Large (PB)}, using triangular membership functions; The fuzzy set of Δe(k): {Rapid Descent (FD), Slow Descent (SD), Stable (ST), Slow Ascent (SU), Rapid Ascent (FU)}, using Gaussian membership functions; Finally, the center-of-gravity (COG) method is used to convert the fuzzy output into precise PID parameters: K i ,K d Similarly where u i is the activation of each rule, The genetic algorithm randomly generates 10 candidate solutions for traffic allocation schemes through population initialization. Each solution includes the traffic proportion of each branch, with the following constraints: ∑Q i = Q total The fitness function is: where P pump is the actual power of the circulation pump, the safety threshold ΔT threshold = 4°C, Select the top 30% of individuals with the highest fitness, and terminate the iteration when the fitness changes less than 1% for 3 consecutive generations, or when the maximum number of iterations (50 generations) is reached. When the temperature of the cooling liquid is lower than the ambient temperature in step Sc1.3, turn on the electric heater to prevent freezing.

9. The adaptive liquid cooling heat dissipation method for land energy storage container according to claim 5, characterized in that: Step Sd is further divided into Sd1, variable frequency circulating pump according to the target total flow Q total Adjust the speed, introduce fuzzy PID output correction; Sd2, the variable orifice regulating valve controls the valve core position through a stepper motor; Sd3, the liquid cooling unit adjusts the compressor power and the expansion valve opening to maintain the cooling liquid temperature; The adjustment accuracy of the variable frequency circulating pump speed in step Sd1 is ±10 rpm; The variable frequency circulating pump speed is determined based on the heat load demand, pipeline impedance, and energy efficiency constraints. The core calculation formula is Wherein, Q total is the total flow demand of the cooling liquid per minute, the rated maximum flow of the pump per minute Q max is determined by the heat load prediction model max N max is the rated maximum rotating speed of the pump; k1, k2 are weight coefficients, k1=0.7, k2=0.3; ΔT tactual is the actual temperature difference between the inlet and outlet of the cooling liquid, ΔT target is the target temperature difference; Fuzzy PID output correction can improve response speed and energy efficiency. The correction formula is N final = N + ΔN PID where e N = N target - N actual , K p , K i , K d are PID parameters generated by the fuzzy rule. The opening resolution of the variable orifice regulating valve in step Sd2 is 0.1 mm; The opening D of the variable orifice regulating valve is dynamically calculated through liquid cooling pipeline flow distribution and impedance matching. The calculation formula is Wherein, Di is the aperture corresponding to the branch valve opening, Q i is the target flow per minute of the corresponding branch, ΔP i is the pressure drop of the corresponding branch, which is obtained in real time by a sensor; C v is the flow coefficient, which is determined by the valve body structure; To solve the coupling interference of multiple branches, pressure balance compensation is introduced D i,final = D i + ΔD i ΔD i = K bal · (P main - P i ) where P main is the main circuit pressure, P i is the corresponding branch inlet pressure, K bal is the balance gain coefficient; To avoid liquid shock in the pipeline, the control commands of the variable frequency pump and the regulating valve need to be calculated synchronously. When the load is high, the circulating pump speed is adjusted first, then the valve opening is optimized, and when the load is low, the valve is closed first, and then the pump speed is reduced. The dynamic decoupling algorithm of this command is: Where α and β are decoupling coefficients, which are updated in real time through pressure feedback; In step Sd3, the liquid cooling unit maintains the cooling liquid temperature at the set value ±5℃.

10. The adaptive liquid cooling heat dissipation method for land energy storage container according to claim 5, characterized in that: Step Se is further divided into Se1.1, flow anomaly detection: actual flow Q of liquid cooling pipeline i If the deviation from the set value lasts for a certain time, it is determined to be a blockage or a leak, the corresponding valve is closed, and the standby redundant pipeline is switched to. Se1.2, temperature consistency detection: if the standard deviation ΔT of the battery cluster temperature is abnormal for a certain period of time, it is determined that the sensor is faulty or locally overheated, the auxiliary cooling scheme is started, and a sensor failure is marked; Se2, alarm and recording, Se2.1, first-level warning, yellow LED light, slight abnormality, system automatic repair; Se2.2, second-level warning, red LED light and buzzer, serious fault, manual intervention required; Se2.3, record fault time, type and processing log to the memory of the main controller module; Actual flow rate Q in step Se1.1 i If the set value is lower than 15% and the state continues for 30 seconds or more, it is determined to be a blockage or a leak. In step Se1.2, if the standard deviation ΔT of the battery cluster temperature is greater than or equal to 3℃ and lasts for more than 1 minute, it is determined that the sensor is faulty or locally overheated; The auxiliary cooling scheme is to start the forced air cooling auxiliary cooling of the adjacent area.

Images