A distributed thermal management system

Abstract

The application discloses a distributed thermal management system, and relates to the technical field of battery thermal management, and realizes efficient temperature equalization of a lithium battery pack through autonomous decision-making of a sub-controller and a consistency protocol between nodes. The system dynamically adjusts PID parameters to adapt to battery aging and environmental changes, and collaborates with global control quantities through lightweight communication, thereby improving response speed by 53% compared with traditional methods, reducing temperature difference by 75%, and reducing energy consumption by 18%.

Description

Technical Field

[0001] This invention relates to the field of battery thermal management technology, and more specifically to a distributed thermal management system. Background Technology

[0002] Existing lithium battery thermal management systems mostly adopt centralized control (such as a single controller managing all battery modules), which has the following problems:

[0003] Single point of failure risk: Failure of the main controller will lead to a complete loss of control;

[0004] High communication latency: Data from multiple nodes needs to be uploaded to the central node for processing, which limits the response speed;

[0005] Poor adaptability: Fixed threshold strategies are difficult to cope with dynamic operating conditions such as battery aging and sudden environmental changes.

[0006] Therefore, it is necessary to propose a distributed thermal management system to solve the above problems. Summary of the Invention

[0007] The purpose of this invention is to solve the problems existing in the background art.

[0008] To achieve the above objectives, the present invention specifically adopts the following technical solution:

[0009] A distributed thermal management system includes multiple battery modules, each module containing several lithium battery cells;

[0010] Each battery module is equipped with an independent controller, and each controller is communicatively connected to the following units:

[0011] The data acquisition unit includes a temperature sensor, a voltage / current sensor, and a battery health status calculation module that are connected to the sub-controller for communication.

[0012] The control unit includes a microcontroller that communicates with the sub-controller and runs an adaptive cooperative control algorithm.

[0013] The communication unit is a CANFD communication module;

[0014] The communication bus, including the CANFD bus, connects all the sub-controllers;

[0015] Each battery module has one heat dissipation actuator.

[0016] Furthermore, the adaptive cooperative control algorithm of the sub-controller includes the following steps:

[0017] Step 1: Distributed data acquisition and preprocessing;

[0018] Step 2: Generation of local adaptive control strategy;

[0019] Step 3: Optimize consistency and collaboration among nodes;

[0020] Step 4: Actuator output and feedback correction.

[0021] Furthermore, the distributed data acquisition and preprocessing in step 1 includes the following steps:

[0022] Step 1.1, Data Input:

[0023] The i-th sub-controller is configured to collect the following data: battery temperature T i Battery voltage V i Current I i Battery health status (SOH) i Ambient temperature T env ;

[0024] Step 1.2, Data Preprocessing:

[0025] Battery temperature T i Perform a moving average filter with a window size of 5 sampling points:

[0026]

[0027] In the formula: T i (k) represents the original temperature value of the i-th module at time k. This is the filtered temperature estimate;

[0028] Normalized cell voltage V i and current I i :

[0029]

[0030] In the formula: i mIn and I max Minimum / maximum allowable current value of battery module, V min and V max Minimum / maximum voltage values ​​allowed for the battery module.

[0031] Furthermore, step 2, the generation of the local adaptive control strategy, includes the following steps:

[0032] Step 2.1, Temperature Deviation Calculation:

[0033] Define target temperature T target =35℃, calculation deviation:

[0034]

[0035] Step 2.2: Dynamically adjust the PID parameters of the actuator:

[0036]

[0037] Where: K p，base =2.0 is the baseline scaling factor, SOH threshol d = 0.8, T env The ambient temperature is represented by 0.2 and 0.1, which are empirical coefficients.

[0038] Step 2.3: Generate local control variables:

[0039] Incremental PID algorithm is used:

[0040] △u i (t)=K p，i ·(e i (t)-e i (t-1))+K i ·e i (t)+K d ·(e i (t)-2e i (t-1)+e i (t-2))

[0041] Where: K p，i SOH is the scaling factor for the i-th node. i The battery health status is represented by a value of 0.8-1.0, T. env For ambient temperature, K i =0.05, K d =0.1.

[0042] Furthermore, step 3, the inter-node consistency and collaborative optimization, includes the following steps:

[0043] Step 3.1, data exchange between neighboring nodes:

[0044] The i-th sub-controller sends to the adjacent node send and the output value u of the control quantity at time t i (t), receiving u j (t);

[0045] Step 3.2, Consistency Protocol Update Control Variables:

[0046] Calculate the global temperature gradient weights:

[0047]

[0048] Update control quantity:

[0049]

[0050] Where γ = 0.2 is the convergence coefficient, which controls the rate of coordinated adjustment. The set of neighbors of node i;

[0051] Step 3.3, Constraint Handling:

[0052] Limit the range of control quantities:

[0053]

[0054] In the formula: Let be the intermediate control quantity calculated collaboratively by the i-th sub-controller at time t through a consensus protocol. The final control quantity of the i-th sub-controller at time t.

[0055] Furthermore, step 4, actuator output and feedback correction, includes the following steps:

[0056] Step 4.1: Map the control quantity to the fan speed:

[0057] Fan speed RPM i Linear relationship with control input:

[0058]

[0059] Step 4.2, Feedback Correction:

[0060] If the actual temperature T i If the deviation between (t+1) and the predicted value exceeds 2℃, parameter adaptation is triggered.

[0061]

[0062] In the formula: 0.05 is the adaptive step size, which balances response speed and stability.

[0063] Compared with the prior art, the beneficial effects of the present invention are as follows:

[0064] 1. The present invention equips each battery module with an independent sub-controller to adjust the heat dissipation strategy in real time based on local data.

[0065] 2. This invention achieves global temperature balance by exchanging status information between nodes through lightweight communication.

[0066] 3. This invention dynamically optimizes control parameters by combining battery state of health (SOH) and environmental parameters.

[0067] 4. Traditional systems rely on a single controller, while this invention makes independent decisions through a sub-controller, so the failure of any node does not affect the whole system.

[0068] 5. Existing solutions use fixed thresholds, while this invention combines battery health status and ambient temperature to adjust parameters in real time, adapting to complex operating conditions.

[0069] 6. Centralized systems require all data to be uploaded to the central node. This invention reduces the bus load by 30% through a lightweight consistency protocol (exchanging only temperature and control values). Attached Figure Description

[0070] Figure 1 This is a flowchart of the main program of the controller of the present invention.

[0071] Figure 2 This is a hardware connection topology diagram of the system of the present invention.

[0072] Figure 3 This is a block diagram of the controller structure of the present invention.

[0073] Figure 4 This is a schematic diagram of the communication protocol fields of the present invention.

[0074] Figure 5 This is the timing diagram for the control signal output of the present invention.

[0075] Figure 6 This is the software task scheduling diagram of the present invention. Detailed Implementation

[0076] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0077] Please see Figures 1 to 6 A distributed thermal management system includes multiple battery modules, each module containing several lithium battery cells.

[0078] Each battery module is equipped with an independent controller, and each controller includes the following units:

[0079] The data acquisition unit includes a temperature sensor (NTC), a voltage / current sensor (Hall effect), and a battery health status calculation module;

[0080] The control unit, a microcontroller (such as STM32H7), runs an adaptive cooperative control algorithm;

[0081] Communication unit, CANFD communication module;

[0082] The communication bus, CANFD bus, connects all sub-controllers, with a transmission cycle of 100ms.

[0083] Actuators, one actuator per battery module, including:

[0084] Cooling fan, brushless DC motor, speed range 1000-3000RPM.

[0085] Specifically, the adaptive cooperative control algorithm of the sub-controller includes the following steps:

[0086] Step 1: Distributed data acquisition and preprocessing;

[0087] Step 1.1, Data Input:

[0088] The i-th sub-controller is configured to collect the following data: battery temperature T i (NTC sensor, sampling frequency 10Hz), battery voltage V i Current I i Battery health status (SOH) i (Calculated using coulomb counting method), ambient temperature T env (SHT30 sensor);

[0089] Step 1.2, Data Preprocessing:

[0090] Battery temperature T i A moving average filter with a window size of 5 sampling points is used. Compared to Kalman filtering, it has lower computational cost and is suitable for embedded real-time systems (such as STM32H7). It reduces random noise from the temperature sensor and smooths data fluctuations.

[0091]

[0092] In the formula: T i (k) represents the original temperature value of the i-th module at time k. This is the filtered temperature estimate;

[0093] Battery temperature noise (such as NTC sensor jitter) can cause control oscillations; filtering improves stability. Normalization provides standardized inputs for subsequent PID control, simplifying parameter tuning.

[0094] Normalized cell voltage V i and current I i : to transfer current I i and voltage V i Normalization to the [0, 1] interval eliminates the difference in voltage / current dimensions, unifies the units, avoids algorithm deviations due to different numerical ranges, and facilitates subsequent control algorithm processing.

[0095]

[0096] In the formula: Imin and I max The minimum / maximum allowable current value of the battery module (e.g., -200A to +200A), V min and V max Minimum / maximum allowable voltage values ​​for the battery module (e.g., 2.5V to 4.2V).

[0097] Step 2: Generation of local adaptive control strategy;

[0098] Step 2.1, Temperature Deviation Calculation:

[0099] Define target temperature T target =35℃, calculation deviation:

[0100]

[0101] Step 2.2: Dynamically adjust PID parameters: Correct the proportional coefficient K in real time based on SOH and ambient temperature. p Adapting to battery aging and environmental changes;

[0102]

[0103] Where: K p,base =2.0 is the baseline scaling factor, SOH threshold =0.8, T env The ambient temperature is represented by 0.2 and 0.1, which are empirical coefficients optimized through experimental calibration. When SOH... threshold When the proportional gain is less than 0.8, the proportional gain increases (aging batteries require stronger control). When the ambient temperature deviates from 25°C, the proportional gain increases with increasing temperature.

[0104] Step 2.3: Generate local control inputs: Fixed PID parameters can lead to insufficient control after battery aging. Dynamic adjustment ensures effectiveness throughout the entire life cycle. The incremental algorithm only outputs changes in control inputs to avoid sudden actuator jumps (such as a fan suddenly running at full speed).

[0105] Incremental PID algorithm is adopted: Incremental PID is suitable for continuous adjustment of actuators (such as fans), avoids integral saturation problem, is suitable for continuous adjustment of fan speed, and reduces overshoot;

[0106] Δu i (t)=K p，i ·(e i (t)-e i (t-1))+K i ·e i (t)+K d ·(e i (t)-2e i (t-1)+e i (t-2));

[0107] Where: K p，i SOH is the scaling factor for the i-th node. i For battery health status (0.8-1.0), T env For ambient temperature, K i =0.05, K d =0.1.

[0108] Step 3: Optimize consistency and collaboration among nodes;

[0109] Step 3.1, data exchange between neighboring nodes:

[0110] The i-th sub-controller sends to the adjacent node send and u i (t), receiving u j (t);

[0111] Step 3.2, Consistency Protocol Updates Control Quantity: Centralized systems rely on a central node, resulting in high communication latency; Distributed consensus protocols achieve global balance through local communication, reducing latency, and the weighting mechanism prioritizes coordinating nodes with similar operating temperatures to avoid the spread of local overheating.

[0112] Calculate the global temperature gradient weights: the smaller the temperature difference between node i and its neighbor node j, the higher the weight w. ij The larger the value, the more likely it is to coordinate with nodes of similar temperature to avoid interference from abnormal nodes.

[0113]

[0114] Update control quantity:

[0115]

[0116] Where γ = 0.2 is the convergence coefficient, which is determined experimentally to balance convergence speed and stability, and to control the rate of coordinated adjustment. The set of neighbors of node i (based on physical proximity or communication topology);

[0117] Step 3.3, Constraint Handling:

[0118] Limit the range of control quantities:

[0119]

[0120] In the formula: Let be the intermediate control quantity calculated by the i-th sub-controller at time t through consensus-based collaborative optimization. The final control quantity of the i-th sub-controller at time t is directly mapped to the operating command (such as speed percentage) of the actuator (e.g., fan, liquid cooling pump). The actuator is driven by a PWM (Pulse Width Modulation) signal. For example: With a 50% PWM duty cycle, the fan speed is 2000 RPM.

[0121] Step 4: Actuator output and feedback correction;

[0122] Step 4.1: Map the control quantity to the fan speed: Map the final control quantity... (Range 0-100) linearly mapped to fan speed (1000-3000RPM). The linear relationship was determined through experiments to ensure that the control quantity 0 corresponds to the lowest speed and 100 corresponds to the highest speed.

[0123] Fan speed RPM i Linear relationship with control quantity: Ensure that the control quantity corresponds to the physical speed range from 0-100%;

[0124]

[0125] Step 4.2, Feedback Correction:

[0126] If the actual temperature T i If the deviation between (t+1) and the predicted value exceeds 2℃, parameter adaptive adjustment is triggered: K is dynamically adjusted according to the actual temperature deviation. p，i Improve robustness;

[0127]

[0128] In the formula: 0.05 is the adaptive step size, which balances response speed and stability;

[0129] If the actual temperature is higher than the prediction (sign = +1), increase K. p，i To enhance control, if the actual temperature is lower than the prediction (sign = -1), reduce K. p，i To avoid over-adjustment.

[0130] Experimental conditions:

[0131] Test subjects: 4 battery modules (each module contains 10 ternary lithium batteries).

[0132] Comparison solution: Traditional centralized PID temperature control system.

[0133] Test conditions: High temperature (40℃ environment), high-rate charge / discharge (2C)

[0134] Experimental results:

[0135] index This invention Traditional centralized system Improvement effect Maximum temperature difference ≤2℃ ≤8℃ Reduced by 75% Response time 1.5 seconds 3.2 seconds shortened by 53% System energy consumption 85W 103W 18% lower Fault recovery time 0.5 seconds System crash No single point of failure

[0136] Tested in a 4-module battery system, compared with a traditional centralized PID controller:

[0137] The maximum temperature difference decreased from 8°C to 2°C;

[0138] The response time has been reduced from 3.2 seconds to 1.5 seconds;

[0139] System energy consumption is reduced by 18%.

[0140] The above are merely preferred embodiments of the present invention and are not intended to limit the present invention. The scope of patent protection of the present invention shall be determined by the claims. Similarly, any equivalent structural changes made based on the description and drawings of the present invention shall also be included within the scope of protection of the present invention.

Claims

1. A distributed thermal management system, characterized in that: comprising a plurality of battery modules, each module containing a number of lithium battery cells; a sub-controller, each battery module is configured with an independent sub-controller, each sub-controller is communicatively connected to the following units: a data acquisition unit, including temperature sensors, voltage / current sensors, and battery health state calculation modules communicatively connected to the sub-controller; a control unit, including a microcontroller communicatively connected to the sub-controller, running an adaptive cooperative control algorithm; a communication unit, which is a CANFD communication module; a communication bus, including a CANFD bus, connecting all sub-controllers; an actuator, each battery module corresponds to a heat dissipation actuator; the adaptive cooperative control algorithm of the sub-controller includes the following steps: Step 1, distributed data acquisition and preprocessing; Step 2, local adaptive control strategy generation; Step 3, inter-node consistency collaborative optimization; Step 4, actuator output and feedback correction.

2. The distributed thermal management system of claim 1, wherein: The distributed data acquisition and preprocessing in step 1 includes the following steps: Step 1.1, data input: The first sub-controller collects the following data: battery temperature , battery voltage , current , battery health status , ambient temperature ; Step 1.2, data preprocessing: On battery temperature A moving average filter is applied with a window size of 5 samples: In the formula: For the first Each module at time The original temperature value, This is the filtered temperature estimate; Normalizing battery voltage and current : In the formulae: and Minimum / maximum current value allowed by the battery module, and Minimum / maximum voltage value allowed by the battery module.

3. A distributed thermal management system according to claim 2, wherein: The step 2 local adaptive control strategy generation includes the following steps: Step 2.1, temperature deviation calculation: Defining target temperature , calculating the deviation: Step 2.2, dynamically adjusting the PID parameters of the actuator: wherein: , is the ambient temperature; 0.2 and 0.1 are empirical coefficients; Step 2.3, generating local control variables: An incremental PID algorithm is used: In the formula: is the proportionality coefficient of the first node, is the state of health of the battery, taking values from 0.8 to 1.0, is the ambient temperature, .

4. The distributed thermal management system of claim 3, wherein: The step 3 inter-node consistency collaborative optimization includes the following steps: Step 3.1, neighbor node data exchange: The first controller sends to the adjacent node and the output value of the control quantity at time receives ​ Step 3.2, consistency protocol update control variable: Calculate the global temperature gradient weight: Update the control variable: wherein, is a convergence coefficient, controlling the rate of the coordinated adjustment, a node a neighbor set of the node; Step 3.3, constraint processing: Limit the control variable range: In the formula: is the th sub-controller at time The intermediate control variable calculated by the collaborative optimization of the consistency protocol, is the th sub-controller at time The final control variable.

5. A distributed thermal management system according to claim 4, wherein: The step 4 actuator is a fan, and its output and feedback correction include the following steps: Step 4.1, mapping the control variable to the fan speed: Fan speed Linear relationship with control quantity: Step 4.2, feedback correction: If the actual temperature If the deviation from the prediction is more than 2°C, the parameter adaptation is triggered: Where: 0.05 is the adaptive step size, balancing response speed and stability.

Images