A dynamic reconfigurable thermoelectric module system and control method of benchmark optimal state design

Abstract

This invention discloses a dynamically reconfigurable thermoelectric module system and control method based on a baseline optimal state design. The system includes a thermoelectric unit array, a programmable switch array, sensor components, a controller, and a heat dissipation structure. Within a feasible domain defined by structural parameters, operating parameters, thermoelectric coupling relationships, and preset constraints, a baseline optimal state (including physical structure, default topology, and operating parameters) is pre-determined for a preset baseline operating condition and user objectives and fixed in hardware. The current heat load is acquired in real time: if it is within the allowable range, the baseline optimal state is maintained; if it exceeds the allowable range, a new optimal state is selected from a pre-stored set or determined online, and the system switches to the corresponding topology and operating parameters; after returning to the allowable range and remaining stable, the system automatically switches back to the baseline optimal state. The heat dissipation structure is configured according to the maximum heat generation power under all reconfiguration modes. This invention anchors the main operating condition with the baseline optimal state and dynamically reconfigures to adapt to changes in operating conditions, achieving high-efficiency energy saving and thermal safety operation of the thermoelectric module throughout its operation.

Description

Technical Field

[0001] This invention relates to the field of semiconductor thermoelectric technology, specifically to a dynamically reconfigurable thermoelectric module system and its control method based on a benchmark optimal state design. It is particularly suitable for application scenarios with dominant operating conditions that require adaptive switching between multiple performance targets, and which also require the module and the peripheral heat dissipation structure to be designed in a coordinated manner, such as thermal management of new energy vehicle batteries, temperature control of communication base stations, heat dissipation of high-performance computing chips, and cooling of precision optical equipment. Background Technology

[0002] Thermoelectric modules operate based on the Peltier effect, offering advantages such as no moving parts, rapid response, and precise temperature control. They are widely used in optoelectronic devices, power battery thermal management, high-performance computing chip heat dissipation, and precision temperature control. As application scenarios become more complex, thermal loads often exhibit dynamic fluctuations. However, many scenarios (such as WLTC conditions in new energy vehicles and periodic loads in communication base stations) have distinct dominant operating conditions—the operating points or ranges that occur most frequently.

[0003] The existing technology has the following main shortcomings:

[0004] First, fixed-topology modules are only compatible with a single design point. Traditional thermoelectric modules use fixed electrical connection topologies (such as all-series or all-parallel connections), and their physical structure is usually designed based on rated operating conditions. When the actual heat load deviates from the design point, the module's coefficient of performance (COP) drops significantly, and this cannot be fundamentally improved by adjusting the current or other means.

[0005] Second, existing optimization methods lack a systematic concept of "optimal state anchor points." Although there is engineering experience in optimizing typical operating conditions, a systematic definition of "optimal state," its generation method, and its anchoring role in control have not yet been formed. Furthermore, there is no technology to use the "optimal state" as a benchmark for system operation and to achieve dynamic switching between multiple optimal states based on this. Existing solutions often treat all operating conditions equally or are designed only for extreme operating conditions, resulting in suboptimal overall performance under the main operating conditions.

[0006] Third, there is a lack of a unified design platform that integrates multi-dimensional technical constraints. Although existing technologies have made partial optimizations to thermoelectric modules from multiple dimensions such as material parameters, structural design, thermo-mechanical coordination, electrical connection, thermal resistance control, spatial layout, reliability protection, and peripheral system adaptation, these optimization measures are mostly aimed at single problems or single operating conditions and are independent of each other. There is a lack of a systematic solution that can be integrated under a unified constraint framework to form a customizable baseline optimal state and support dynamic operation.

[0007] Fourth, the module and its external heat dissipation structure lack coordinated design. Existing optimization schemes for thermoelectric modules only focus on the internal structure of the module, neglecting the adaptation of the external heat dissipation structure. When the module's operating mode changes, the heat output at its hot end can change several times. If the heat dissipation structure is configured only for the main operating condition, the hot end temperature will rise sharply under extreme conditions; if it is configured for extreme conditions, it will result in wasted costs under the main operating condition.

[0008] Therefore, there is an urgent need in this field for a complete system and its control method that can integrate multi-dimensional technical constraints, generate customizable benchmark optimal states, realize dynamic reconfiguration of the thermoelectric module, and coordinate with the design of the external heat dissipation structure, so as to simultaneously meet the comprehensive requirements of optimal overall performance under main operating conditions, availability under extreme operating conditions, controllable switching frequency, and thermal safety assurance. Summary of the Invention

[0009] To address the above problems, this invention proposes a dynamically reconfigurable thermoelectric module system and control method based on a baseline optimal state design. The core of this invention lies in: within a feasible domain defined by structural parameters, operating parameters, and thermoelectric coupling relationships, and constrained by preset constraints, a "baseline optimal state" is pre-determined based on historical heat load statistics and user objectives and solidified into hardware. Using this baseline optimal state as an anchor point, dynamic reconfiguration is performed between multiple optimal states through a restricted triggering strategy. Simultaneously, the peripheral heat dissipation structure is designed collaboratively with the module to ensure that the hot-end temperature remains within a safe range, achieving internal and external coordination between the module and the peripheral system.

[0010] This invention integrates "offline optimization and solidification + online controlled reconstruction + peripheral collaborative configuration" into a unified operating mechanism of "statistics → optimization → solidification → deviation judgment → reconstruction → regression → hot end security assurance".

[0011] The specific technical solution is as follows:

[0012] In a first aspect, the present invention provides a dynamically reconfigurable thermoelectric module system based on a benchmark optimal state design, comprising:

[0013] Thermoelectric unit array, composed of multiple P-type and N-type semiconductor legs;

[0014] Programmable switch arrays are used to change the electrical connection topology between thermoelectric units;

[0015] Sensor components are used to acquire data related to heat load;

[0016] The controller is communicatively connected to the sensor assembly and the programmable switch array;

[0017] A heat dissipation structure is coupled to the thermal side of the thermoelectric module to dissipate the heat generated by the module to the external environment.

[0018] Figure 1 The diagram shows the structure of the system of the present invention. As shown in the figure, the thermoelectric unit array (1) is connected to the programmable switch array (2). The sensor component (3) collects the heat load data of the thermoelectric unit array (1) and transmits it to the controller (4). The controller (4) controls the programmable switch array (2) and the adjustable power supply module (6) according to the preset logic. The heat dissipation structure (5) is coupled to the thermal side of the module.

[0019] The preset baseline operating condition is a representative operating condition determined based on historical heat load statistics, including heat load values ​​and their spatial distribution characteristics. The spatial distribution characteristics of the heat load are used to characterize the power density distribution, hot spot location, and heat concentration of the heat source within the operating area of ​​the thermoelectric module, so that the generated baseline optimal state can adapt to the actual thermal field characteristics of the target application scenario, rather than being optimized solely based on the total heat load value.

[0020] Within the feasible domain, which is composed of structural parameters, operating parameters, and thermoelectric coupling relationships, and is limited by preset constraints, the optimal performance configuration for a given heat load condition and preset performance target is defined as the optimal state. The optimal state includes the corresponding physical structural parameters, electrical connection topology, and operating parameters.

[0021] The physical structure parameters and default electrical connection topology of the thermoelectric unit array are the optimal state determined in advance within the feasible domain based on the preset benchmark operating conditions determined by historical heat load statistics and the user's preset performance targets. This is called the benchmark optimal state and is fixed into the hardware configuration.

[0022] The controller is configured to:

[0023] (1) The baseline optimal state is taken as the default operating state of the system;

[0024] (2) By real-time monitoring of the deviation between the current heat load and the preset reference operating condition, or by judging whether the current operating condition is within the allowable range corresponding to the preset reference operating condition according to the preset rules.

[0025] (3) If it is within the allowable range, the baseline optimal state is maintained; if it exceeds the allowable range, a new optimal state applicable to the current working condition is obtained (selected from the pre-stored optimal state set or determined online), the programmable switch array is controlled to switch to the electrical connection topology corresponding to the new optimal state, and the operating parameters corresponding to the new optimal state are applied;

[0026] (4) When the current working condition is re-determined to be within the allowable range corresponding to the preset benchmark working condition and remains stable for a period of time, the system automatically switches back to the benchmark optimal state.

[0027] Figure 2The diagram shows the overall architecture and control flow of the system of this invention, illustrating the complete closed loop of the offline design phase (statistical heat load distribution → determining the baseline operating condition → generating the baseline optimal state within the feasible domain → solidifying it into hardware and configuring heat dissipation) and the online operation phase (default operation → deviation / preset law judgment → maintenance or switching → regression).

[0028] Furthermore, the heat dissipation capacity of the heat dissipation structure is configured according to the maximum heat generation power of the module in all reconfiguration modes, so that the hot end temperature of the module does not exceed a preset safety threshold in any reconfiguration mode.

[0029] Furthermore, the deviation is obtained by calculating the deviation rate δ between the current heat load and the preset reference condition, where the deviation rate is the absolute value of the deviation, i.e., δ = |current heat load - reference condition| / reference condition. The allowable range is determined by a preset deviation threshold. Limited; when δ≤ When it is determined to be within the allowable range, when δ> The deviation threshold is determined to be exceeded. Preferably, the deviation threshold is... = k· / ,in The heat load value corresponding to the preset benchmark operating condition. The standard deviation of the heat load is k, which is 2 to 3; the duration of steady-state time is... = n·τ, where n ≥ 3, and τ is the thermal time constant of the thermoelectric module. See also Figure 3 As shown. The determination of the above deviation threshold and settling time follows the principles below:

[0030] Fluctuations in heat load exhibit statistical characteristics, with a standard deviation of [missing information]. This reflects the degree of dispersion of the heat load relative to the mean. Deviation threshold. = k· / This means that the allowable range is set as a statistical interval near the baseline operating condition, and the value of k determines the width of this interval. When k=2, the allowable range is approximately the mean ±2. (Approximately 95% confidence interval); when k=3, it is approximately the mean ± 3. (Approximately 99.7% confidence interval). Engineering values ​​of 2-3 avoid frequent unnecessary topology switching due to normal fluctuations while also ensuring timely response when operating conditions deviate significantly. Settling time. = n·τ (n≥3) is set based on the thermal time constant τ of the thermoelectric module. The step response of a first-order thermal system reaches 95% of the steady-state value after 3 times the time constant. Therefore, taking n≥3 can ensure that the system has fully responded and tended to a new thermal equilibrium state before performing regression, avoiding frequent back-and-forth switching caused by temporary fluctuations.

[0031] In a preferred embodiment (see Example 1), WLTC cycle testing showed an average COP of 0.51, with approximately 6 switching actions per hour, a reduction of 86% compared to the fully dynamic solution. The battery temperature remained ≤45°C, while the average COP of the traditional fixed topology module was only 0.28 under the same operating conditions. The average COP of the present invention is improved by 82%.

[0032] Furthermore, in the preset rule-triggered mode, the controller does not rely on real-time heat load deviation calculations, but directly triggers topology reconfiguration based on pre-determined rules. These preset rules include, but are not limited to:

[0033] (1) Preset time series: Triggered at fixed or variable time intervals, such as switching the topology every 10 minutes. Figure 6 An example of a periodic reconstruction mode based on a preset time series is given, in which the controller directly switches the topology according to the preset schedule without the need to calculate the deviation in real time;

[0034] (2) Operating condition stage table: triggered according to the working stage of the equipment (such as discharging, resting, charging), for example, the three-stage process of the capacity-forming equipment;

[0035] (3) Event triggering rules: Triggered by external events (such as temperature alarm signals, power change commands), for example, immediately switching to high cooling capacity topology after receiving an over-temperature protection signal.

[0036] Because the operating condition changes under the preset triggering mode have known periodicity or predictability, the controller does not need to solve for a new optimal state online. Instead, it directly selects a new optimal state corresponding to the current operating condition from the offline pre-stored set of optimal states, thereby significantly reducing computational intensity and response latency. The pre-stored set of optimal states is pre-stored after optimization for various possible operating conditions (such as low load, medium load, and high load) during the offline design phase.

[0037] In a preferred embodiment (see Example 2), for the three-stage fixed charge and discharge program (discharge → rest → charge) of the battery formation and capacity testing equipment, the controller directly selects the corresponding topology from the pre-stored optimal state set according to the preset time sequence, without the need to calculate the deviation rate in real time. The sensor sampling frequency can be reduced from 10Hz to 0.1Hz, the computing load is reduced by 90%, and the hardware cost is reduced by about 12%.

[0038] Furthermore, in this invention, the preset constraints are used to limit the allowable combination range between the structural parameters, operating parameters, and peripheral system conditions of the thermoelectric module, thereby forming a constrained feasible region that satisfies both engineering feasibility and operational stability. The baseline optimal state is not an unconstrained optimization result, but rather the optimal state in terms of overall performance within the feasible region for the target operating condition. For example, the constraints may originate from one or more of the following: material properties, electrothermal coupling relationships, thermodynamic behavior, thermal management requirements, reliability requirements, peripheral system adaptation, and manufacturing implementation conditions. Those skilled in the art can select applicable constraints according to actual application scenarios, all of which fall within the scope of protection of this invention.

[0039] Furthermore, the user-preset performance objectives include at least one of the following: maximizing energy efficiency ratio, maximizing cooling capacity, minimizing cost, minimizing temperature difference, achieving optimal cost-effectiveness, or any user-defined objective function. For example... Figure 4 As shown, different user objectives (such as energy efficiency priority, temperature balance, cooling priority, and extreme cooling) correspond to different baseline optimal states on the Pareto front, and users can choose the appropriate operating point according to their needs.

[0040] In a preferred embodiment (see Example 3), for high-performance computing chip heat dissipation scenarios, the user aims to maximize cooling capacity and determines offline... =125W, optimized to the default topology of 2 series 8 parallel, with a cooling capacity of 80W at 125W. When running online, when the CPU thermal load suddenly increases to 200W, it switches to 1 series 16 parallel, providing 110W of cooling capacity, keeping the chip from throttling, and the hot end temperature is ≤75℃ throughout.

[0041] Furthermore, the controller is configured to execute a controlled switching mechanism before switching the electrical connection topology. This controlled switching mechanism includes at least one of soft switching, current-limiting switching, and phased switching. During topology switching, direct operation of the switch may generate voltage spikes, current surges, or arcing, affecting the lifespan of the switching devices and system reliability. In addition, for thermoelectric modules, the reversal of the current direction during switching may cause a change in the Peltier effect, turning the cold end into the hot end, directly affecting the normal operation of the controlled device. Therefore, this invention introduces a controlled switching mechanism to proactively intervene in the electrical state before switching the electrical connection topology, thereby reducing electrical stress during the switching process and improving system reliability. Soft switching is achieved by first reducing the operating current to a safe level before operating the switch; current-limiting switching avoids abrupt changes by limiting the rate of current change; and phased switching breaks down the switching process into multiple steps for gradual transition. These mechanisms can be used individually or in combination to reduce electrical stress during the switching process and ensure the safe and reliable operation of the switching devices and modules. Figure 5The timing diagram of soft switching is shown, which includes the current reduction stage, the switching action stage (turning on first and then turning off), and the current recovery stage, to ensure a smooth and safe switching process.

[0042] In a preferred embodiment (see Example 4), the controller reduces the operating current to a safe level before switching topologies and restores the current to the target value after the switching action is completed. If the current direction of the target topology is opposite to the current, the current is first reduced to zero, and then established in the new direction after the switching is completed, avoiding instantaneous reversal of hot and cold ends. Testing has shown that this mechanism ensures uninterrupted current and no voltage spikes during the switching process, ensuring the safety and reliability of the switching devices.

[0043] In a second aspect, the present invention provides an adaptive control method based on the above-mentioned system, comprising the following steps:

[0044] S1 performs offline statistical analysis of the heat load distribution of the target application scenario, determines the preset baseline operating conditions, and generates the baseline optimal state within the feasible domain defined by the user's preset performance targets and preset constraints, including physical structural parameters and default electrical connection topology, and solidifies it into hardware.

[0045] S2: When running online, the default running state is the baseline optimal state;

[0046] S3: By real-time monitoring of the deviation between the current heat load and the preset reference operating condition, or according to preset rules, determine whether the current operating condition is within the allowable range corresponding to the preset reference operating condition.

[0047] S4: If within the allowable range, maintain the baseline optimal state; if outside the allowable range, obtain a new optimal state applicable to the current operating condition (selected from the pre-stored optimal state set or determined online), control the programmable switch array to switch to the electrical connection topology corresponding to the new optimal state, and apply the operating parameters corresponding to the new optimal state;

[0048] S5: When the current working condition is re-determined to be within the allowable range corresponding to the preset benchmark working condition and remains stable for a certain period of time, it will automatically revert to the benchmark optimal state.

[0049] Furthermore, S1 also includes: configuring the external heat dissipation structure according to the maximum heat generation power of the module in all reconfiguration modes to ensure that the hot end temperature is always safe.

[0050] Furthermore, the preset rules include at least one of preset time series, working condition stage table, and event triggering rules; when topology reconstruction is directly triggered according to the preset rules, the corresponding new optimal state is selected from the pre-stored optimal state set.

[0051] Furthermore, a controlled handover mechanism is executed before the topology reconstruction, and the controlled handover mechanism includes at least one of soft handover, rate-limiting handover, and phased handover.

[0052] Thirdly, the present invention provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the above-described method.

[0053] Fourthly, the present invention provides an electronic device comprising a processor and a memory, wherein the processor implements the above-described method when executing instructions in the memory.

[0054] Beneficial effects

[0055] 1. A "benchmark optimal state" anchoring mechanism is proposed: A benchmark optimal state, pre-optimized and solidified into hardware for the main operating condition, serves as a fixed reference point for system operation, forming a complete control closed loop of "offline solidified anchor point → online deviation monitoring → controlled reconfiguration → automatic regression." This mechanism unifies module design and control within the same framework, enabling multi-dimensional design constraints to work synergistically under this unified framework.

[0056] 2. Optimal overall performance under main operating conditions: Ensures that the module is in a state of optimal overall performance under the most frequent operating conditions, significantly improving actual operating energy efficiency.

[0057] 3. Significantly reduce switching losses: such as Figure 7 As shown, the number of switching actions is significantly reduced through the constrained reconfiguration and automatic regression mechanism, thereby significantly reducing switching losses and electromagnetic interference.

[0058] 4. Controllable hardware costs: The power supply and switching capacity are designed based on the main operating conditions rather than extreme operating conditions. Extreme operating conditions are achieved through topology reconfiguration without increasing hardware capacity.

[0059] 5. Diverse triggering methods: It supports both closed-loop control based on real-time deviation and open-loop control based on preset rules, making it suitable for periodic load scenarios.

[0060] 6. Module-Heat Dissipation Coordination: The heat dissipation structure is configured according to the worst-case operating conditions to ensure that the hot end temperature of the module does not exceed the safe threshold in any reconfiguration mode, thus avoiding performance crashes caused by insufficient heat dissipation.

[0061] 7. Unified Platform Framework: This invention unifies the design and operation of modules under a framework of "benchmark optimal state anchoring + constrained reconstruction + automatic regression." This framework does not depend on any specific type of constraint or performance target and is compatible with different application scenarios and design requirements. This platform characteristic not only enables it to integrate current multi-dimensional design constraints but also provides an infrastructure for subsequent technology iterations and expanded applications, exhibiting excellent scalability. Attached Figure Description

[0062] Figure 1 This is a schematic diagram of the system structure of the present invention;

[0063] Figure 2 This is a diagram illustrating the overall architecture and control flow of the system of the present invention.

[0064] Figure 3 A schematic diagram of the probability distribution of heat load and a method for determining the baseline operating condition;

[0065] Figure 4 Illustration of Pareto frontier and baseline state selection for different user objectives;

[0066] Figure 5 This is a timing diagram of a controlled handover (soft handover).

[0067] Figure 6 This is a schematic diagram of a periodic reconstruction pattern based on a preset time series.

[0068] Figure 7 A comparison chart of the number of switching actions with the fully dynamic reconfiguration scheme;

[0069] Explanation of reference numerals in the attached figures:

[0070] (1) Thermoelectric unit array (2) Programmable switch array (3) Sensor assembly (4) Controller (5) Heat dissipation structure (6) Adjustable power supply module Detailed Implementation

[0071] The present invention will now be described in detail with reference to the embodiments. These embodiments are for illustrative purposes only and do not constitute a limitation on the scope of protection thereof.

[0072] Example 1: On-board battery thermal management module (deviation trigger mode)

[0073] A new energy vehicle manufacturer provided WLTC operating condition data and requested a customized thermal management module, prioritizing energy efficiency while also considering reliability.

[0074] Offline phase: Statistical heat load probability distribution (see...) Figure 3 The frequency of occurrence in the 50W-60W range is 75%, and the heat load value corresponding to the preset reference operating condition is set. =55W. The user's goal is to "maximize the coefficient of performance (COP)". The feasible region is defined by preset constraints (including material matching constraints, thermodynamic synergistic constraints, thermal resistance and thermal management constraints, etc.), and is determined by structural parameters ( Within the feasible region comprised of factors such as leg height, substrate thickness, default topology, operating parameters (operating current), and thermoelectric coupling, the optimal physical parameters are determined with the objective of maximizing the COP under 55W operating conditions. =1.2, leg height 1.5mm, substrate thickness 1.5mm, default topology 4 series 4 parallel, obtained the baseline optimal state S*, COP=0.52, and fixed as hardware. The controller pre-stores the reconfigurable topology library (8 series 2 parallel for low load, 2 series 8 parallel for high load).

[0075] The heat dissipation structure is configured in a coordinated manner: Under high load, the equivalent resistance of the 2-series 8-parallel topology is approximately 0.22Ω, and the Joule heat at 15A is approximately 50W, plus Peltier heat, resulting in a total heat generation of approximately 200W. With this liquid cooling radiator configuration, the thermal resistance is 0.05K / W, and the hot-end temperature at an ambient temperature of 30℃ is 30 + 0.05 × 200 = 40℃, far below the safe upper limit of 80℃. If the radiator is configured only for the main operating condition of 55W, the hot-end temperature will exceed 100℃ under high load, leading to a collapse in cooling efficiency. The internal and external coordinated design of this invention ensures thermal safety under all operating conditions.

[0076] Deviation Trigger Mode: Deviation Threshold =0.1 (calculated from the standard deviation of heat load), thermal time constant τ=38s, settling time =120s.

[0077] Online operation: WLTC cycle test shows an average COP of 0.51, with approximately 6 switching actions per hour, a reduction of 86% compared to the fully dynamic solution. The battery temperature is consistently ≤45℃, and the hot end temperature is consistently ≤50℃.

[0078] In comparison, the traditional fixed topology module (fully series-connected, optimized for 55W) has an average COP of only 0.28 under the same operating conditions, with battery temperature fluctuations reaching ±8℃, exceeding 52℃ during high-load periods. Exemplary results show that the present invention improves the average COP by 82% and stabilizes the battery temperature below 45℃.

[0079] Example 2: Periodic Reconstruction Pattern Based on Preset Time Series

[0080] A battery formation and capacity testing device has a fixed three-stage charging and discharging program: 0-10 minutes of constant current discharge (high load, approximately 150W), 10-20 minutes of rest (low load, approximately 50W), and 20-30 minutes of constant voltage charging (medium load, approximately 90W), repeating this process daily. In this embodiment, the controller pre-stores this time-load mapping table and directly selects the corresponding topology from the pre-stored optimal state set according to the preset time sequence (discharging → low resistance parallel connection, rest → high resistance series connection, charging → medium resistance hybrid connection), without needing to calculate the deviation rate in real time. Figure 6 As shown, the heat dissipation structure is configured according to the maximum heat generation power during the discharge phase. Tests show that the sensor sampling frequency can be reduced from 10Hz to 0.1Hz, the controller's computational load is reduced by 90%, the hardware cost is reduced by about 12%, and the cooling performance fully meets the process requirements.

[0081] Example 3: Heat dissipation of high-performance computing chips (cooling capacity priority)

[0082] The server CPU operates at 120W-130W, and the user requires maximum cooling capacity. (Offline determination needed.) =125W, optimized to a default topology of 2 series and 8 parallel, with an equivalent resistance of 0.15Ω, resulting in a cooling capacity of 80W at 125W. Figure 4 As shown by point ③, this baseline optimal state is located in the cooling-priority region of the Pareto front. The heat dissipation structure is configured according to the maximum heat dissipation power (approximately 400W) under a 1-series 16-parallel topology, employing a large-size heatsink and a high-speed fan. During online operation, when the CPU thermal load suddenly increases to 200W, it switches to a 1-series 16-parallel topology, providing 110W of cooling capacity to prevent the chip from throttling; upon regression, it automatically switches back to the baseline optimal state. The hot-end temperature remains ≤75℃ throughout the entire process.

[0083] Example 4: Controlled handover mechanism

[0084] This embodiment illustrates the controlled switching mechanism performed by the controller before topology reconfiguration. Before switching the topology, the controller reduces the operating current to a safe level (e.g., below 10% of the rated value), completes the switching action, and then restores the current to the target value, such as... Figure 5 As shown, during the switching process, the controller also ensures the consistency of the current direction before and after the switch: if the current direction of the target topology is opposite to the current, the current is first reduced to zero during the switching process, and then established in the new direction after the switch is completed, thereby avoiding instantaneous reversal of the hot and cold ends. Specific implementations can employ different methods such as soft switching (turning on first and then off), current-limiting switching (limiting the rate of change of current), or staged switching. Those skilled in the art can select appropriate switching strategies based on the actual power supply voltage and the specifications of the switching devices; all of these are equivalent embodiments of the present invention. Testing has shown that this mechanism ensures uninterrupted current and no voltage spikes during the switching process, ensuring the safety and reliability of the switching devices.

Claims

1. A dynamically reconfigurable thermoelectric module system based on a baseline optimal state design, characterized in that, include: Thermoelectric unit array; Programmable switch arrays are used to change the electrical connection topology between thermoelectric units; Sensor components are used to acquire heat load data; The controller is connected to the sensor assembly and the programmable switch array; Among them, the preset benchmark operating condition is a representative operating condition that is determined based on historical heat load statistics and includes heat load values ​​and spatial distribution characteristics; Within the feasible domain, which is composed of structural parameters, operating parameters, and thermoelectric coupling relationships, and is limited by preset constraints, the optimal performance configuration determined for a given thermal load condition and preset performance target is defined as the optimal state. The optimal state includes the corresponding physical structural parameters, electrical connection topology, and operating parameters. The physical structure parameters and default electrical connection topology of the thermoelectric unit array are the optimal state determined in advance within the feasible domain based on the preset benchmark operating conditions and user preset performance targets. This optimal state is called the benchmark optimal state and is fixed as hardware configuration. The controller is configured to: (1) use the benchmark optimal state as the default operating state of the system; (2) determine whether the current operating condition is within the allowable range corresponding to the preset benchmark operating condition by real-time monitoring of the deviation between the current heat load and the preset benchmark operating condition, or according to the preset rules. (3) If it is within the allowable range, the baseline optimal state is maintained; if it exceeds the allowable range, a new optimal state applicable to the current working condition is obtained (selected from the pre-stored optimal state set or determined online), the programmable switch array is controlled to switch to the electrical connection topology corresponding to the new optimal state, and the operating parameters corresponding to the new optimal state are applied; (4) When the current working condition is re-determined to be within the allowable range corresponding to the preset benchmark working condition and remains stable for a period of time, the system automatically switches back to the benchmark optimal state.

2. The system according to claim 1, characterized in that, It also includes a heat dissipation structure coupled to the thermal side of the thermoelectric module; the heat dissipation capacity of the heat dissipation structure is configured according to the maximum heat generation power of the module in all reconfiguration modes, so that the hot end temperature of the module does not exceed a preset safety threshold in any reconfiguration mode.

3. The system according to claim 1, characterized in that, The deviation is obtained by calculating the deviation rate δ between the current heat load and the preset baseline operating condition, and the allowable range is determined by a preset deviation threshold. Limited; when δ≤ When it is determined to be within the allowable range, when δ> The time limit is exceeded.

4. The system according to claim 3, characterized in that, The deviation threshold = k· / ,in The heat load value corresponding to the preset benchmark operating condition. The standard deviation of the heat load is k, which is 2 to 3; the duration of steady-state time is... = n·τ, where n ≥ 3, and τ is the thermal time constant of the thermoelectric module.

5. The system according to claim 1, characterized in that, The preset rules include at least one of preset time series, operating condition stage table, and event triggering rules. When topology reconstruction is directly triggered according to the preset rules, the corresponding new optimal state is selected from the pre-stored optimal state set.

6. The system according to claim 1, characterized in that, The preset constraints are used to limit the allowable combination range between the structural parameters, operating parameters and peripheral system conditions of the thermoelectric module, so as to form a constrained feasible domain, so that the benchmark optimal state meets the requirements of engineering feasibility, operational stability and target operating condition adaptation.

7. The system according to claim 1, characterized in that, The user-preset performance objectives include at least one of the following: maximizing energy efficiency ratio, maximizing cooling capacity, minimizing cost, minimizing temperature difference, achieving optimal cost-effectiveness, or any user-defined objective function.

8. The system according to claim 1, characterized in that, The controller is also configured to perform a controlled switching mechanism before switching electrical connection topologies, the controlled switching mechanism including at least one of soft switching, current-limiting switching, and phased switching.

9. An adaptive control method based on the system according to any one of claims 1 to 8, characterized in that, Includes the following steps: S1: Offline statistical analysis of the heat load distribution of the target application scenario, determine the preset benchmark operating conditions, and generate the benchmark optimal state within the feasible domain, which is composed of structural parameters, operating parameters and thermoelectric coupling relationship and is limited by preset constraints, according to the user's preset performance target and preset constraints. This includes physical structural parameters and default electrical connection topology, and is then fixed into hardware. S2: When running online, the default running state is the baseline optimal state; S3: By real-time monitoring of the deviation between the current heat load and the preset reference operating condition, or according to preset rules, determine whether the current operating condition is within the allowable range corresponding to the preset reference operating condition. S4: If it is within the allowable range, then maintain the baseline optimal state; If the allowable range is exceeded, a new optimal state applicable to the current operating condition is obtained (selected from the pre-stored optimal state set or determined online), the programmable switch array is controlled to switch to the electrical connection topology corresponding to the new optimal state, and the operating parameters corresponding to the new optimal state are applied; S5: When the current working condition is re-determined to be within the allowable range corresponding to the preset benchmark working condition and remains stable for a certain period of time, it will automatically revert to the benchmark optimal state.

10. The method according to claim 9, characterized in that, S1 also includes: configuring an external heat dissipation structure based on the module's maximum heat generation power in all reconfiguration modes to ensure that the hot end temperature is always safe.

11. The method according to claim 9, characterized in that, The preset rules include at least one of preset time series, working condition stage table, and event triggering rules; when topology reconstruction is directly triggered according to the preset rules, the corresponding new optimal state is selected from the pre-stored optimal state set.

12. The method according to claim 9, characterized in that, Before the topology reconstruction, a controlled handover mechanism is executed, which includes at least one of soft handover, rate-limiting handover, and phased handover.

13. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by a processor, it implements the method of any one of claims 9 to 12.

14. An electronic device comprising a processor and a memory, characterized in that, The processor implements the method of any one of claims 9 to 12 when executing instructions in the memory.

Images