Composite phase change energy storage and thermal management system and optimization method thereof

Abstract

The application discloses a composite phase change energy storage and heat management system and an optimization method thereof. The system comprises a TPMS metal support structure module, a phase change energy storage module, a temperature field monitoring module, a field driving optimization control module, a heat flow regulation module and an energy recovery module. By introducing a temperature field driving design strategy, a nonlinear mapping relationship between the temperature field and the porosity distribution is established by using nTopology software, a multi-scale porosity gradient continuously changing from low to high is formed, and a dynamic balance between the heat conduction performance and the energy storage capacity is realized. The system obtains the internal temperature distribution in real time through the sensing module, calculates the optimal porosity distribution and the heat flow path through the field driving algorithm, and realizes the adaptive adjustment of the heat performance. Compared with the traditional uniform porosity metal foam composite material, the application significantly improves the heat conduction efficiency and the melting rate, shortens the phase change time by more than 20%, and increases the total energy storage capacity by about 10%. The application can be widely applied to electronic device cooling, solar heat storage, heat management and intelligent energy systems.

Description

Technical Field

[0001] This invention relates to the field of energy storage and thermal management technology, and in particular to a triple-period minimal surface structure composite phase change energy storage and thermal management system based on field-driven design and its optimization method. Background Technology

[0002] In recent years, the global energy structure has been rapidly transforming towards low-carbon and high-efficiency utilization, and thermal energy storage and intelligent thermal management technologies have become important supports for improving the efficiency of energy systems. Phase change material (PCM) energy storage, due to its high latent heat density, constant temperature characteristics, and recyclability, has been widely used in solar thermal utilization, building energy conservation, electric vehicles, and electronic device cooling. However, traditional PCM materials suffer from problems such as low thermal conductivity, slow phase change rate, and insufficient thermal energy utilization, leading to sluggish system response and reduced thermal efficiency.

[0003] To improve thermal conductivity, researchers often combine PCM with metal foam, fins, or porous materials. However, such structures still face the following technical bottlenecks: fixed porosity and lack of targeted design; the pore structure of traditional metal foam materials is uniformly distributed, making it difficult to optimize the heat conduction path based on temperature gradient or heat flow direction; structural parameter optimization relies on experience, and existing studies mostly adopt linear or piecewise gradient porosity designs, which cannot fully match the actual nonlinear temperature field variation.

[0004] Triple Periodic Minimal Surface (TPMS) structures have attracted widespread attention due to their continuous thermal conductivity, high specific surface area, and excellent mechanical stability. Additive manufacturing technology allows for the high-precision fabrication of complex TPMS topologies, demonstrating their potential in phase change energy storage, lightweight heat sinks, and heat recovery. Existing research indicates that TPMS structures can improve thermal conductivity by 15%-30% compared to traditional metal foams and effectively reduce temperature non-uniformity.

[0005] However, current research on TPMS-PCM composite structures, both domestically and internationally, is still focused on static structural design and finite gradient optimization, and has not yet solved the following key problems: it is impossible to achieve nonlinear porosity changes that match the actual temperature field; and it lacks a dynamic optimization control mechanism that combines real-time data feedback and intelligent algorithms.

[0006] In summary, there is an urgent need for a TPMS composite phase change system capable of: driving structural parameter changes based on temperature field information; achieving intelligent control of nonlinear porosity distribution; and combining reinforcement learning algorithms to achieve adaptive thermal performance optimization. This invention addresses these issues by proposing a field-driven design-based triple-period minimal surface structure composite phase change energy storage and thermal management system and its optimization method, demonstrating innovation and application prospects in theory, design, manufacturing, and control.

[0007] The differences compared to existing technologies are as follows:

[0008] Technical Comparison with Patent CN120016949B "Photovoltaic Energy Storage and Cooling Phase Coupling System and Method Based on Copper-based Composite Phase Change Material"

[0009] Patent CN120016949B constructs a coupled energy system for photovoltaic power generation. Its core consists of a photovoltaic power generation module, a copper-based composite phase change material cooling and energy storage module, a thermoelectric conversion module, and an energy recovery system, primarily addressing the problems of reduced power generation efficiency due to excessively high photovoltaic panel temperatures and the inability to generate electricity at night. In contrast, this patent constructs a general-purpose multi-field coupled thermal management system model. By introducing a triple-period minimal surface (TPMS) metal support structure and a phase change material composite, combined with a field-driven algorithm, it achieves dynamic adjustment of the heat conduction path and energy storage capacity, applicable to various scenarios such as electronic device cooling, energy storage, and heat recovery. The two differ significantly in their system structure design and application direction.

[0010] Patent CN120016949B enhances thermal conductivity at the material level by employing a high thermal conductivity copper-based composite phase change material (thermal conductivity approximately 400 W / m·K). It utilizes physical composite and electroplating modification to improve interfacial thermal conductivity, achieving rapid heat transfer and energy storage release. This patent, however, employs a structure-field synergistic optimization mechanism. Based on a temperature-field-driven nonlinear porosity mapping relationship, it constructs a continuously varying TPMS multi-scale structural gradient from low to high, achieving a dynamic balance between enhanced thermal conductivity in high-temperature regions and enhanced energy storage in low-temperature regions. The former belongs to a material composite thermal conductivity enhancement mechanism, while the latter belongs to a structure-functional gradient and field-driven synergistic optimization mechanism; their thermal conductivity regulation principles and implementation paths are fundamentally different.

[0011] Patent CN120016949B employs an energy operation control method based on a deep reinforcement learning (DDPG) algorithm. By defining a state space, action space, and reward function, it optimizes heat dissipation intensity, thermoelectric energy conversion efficiency, and electrical energy output, achieving intelligent scheduling and efficient energy management of photovoltaic systems. This patent further expands upon this approach into a field-driven intelligent optimization control system. It embeds the deep reinforcement learning algorithm into the entire process of structural design and operation, constructs a dynamic optimization model of porosity distribution driven by a temperature field, and combines a surrogate model and nonlinear mapping to achieve real-time porosity adjustment. The two patents differ in their algorithm application levels; the former focuses on control optimization at the system operation level, while the latter achieves intelligent optimization throughout the entire process from structural design to operational adjustment.

[0012] Patent CN120016949B employs traditional manufacturing methods such as mechanical crushing, electroplating, mixing, casting, and heat treatment, suitable for mass production and engineering integration, focusing on improving the thermal conductivity and structural stability of composite materials. This patent, however, utilizes an additive manufacturing process (3D printing) driven by implicit equations. It generates multi-scale TPMS implicit function models through the nTopology design platform and combines vacuum impregnation and metal surface heat treatment technologies to achieve high-precision structural forming. This method can flexibly adjust the structural porosity according to the thermal field distribution, achieving closed-loop optimization from virtual model design to physical manufacturing, combining intelligent design and manufacturing adaptability. The two differ significantly in their manufacturing paths, manufacturing precision, and degrees of freedom in structural control: the former leans towards the feasibility of traditional composite processes, while the latter reflects the new trend of digital and intelligent manufacturing. Summary of the Invention

[0013] This invention aims to overcome the shortcomings of traditional phase change energy storage systems, such as low thermal conductivity, fixed structural parameters, and inability to achieve adaptive control. It provides a field-driven design-based triple periodic minimal surface (TPMS) structure composite phase change energy storage and thermal management system and its optimization method. This invention achieves a dynamic balance between thermal performance and energy storage capacity by establishing a nonlinear mapping relationship between the temperature field and porosity distribution. By combining reinforcement learning algorithms and field-driven control mechanisms, it achieves intelligent self-optimization of structure and thermal performance, thereby significantly improving the thermal response speed and energy utilization rate of the phase change energy storage system.

[0014] To achieve the above objectives, the technical solution adopted by the present invention is as follows:

[0015] The composite phase change energy storage and thermal management system of the present invention mainly consists of a TPMS metal support structure module, a phase change energy storage module, a temperature field monitoring module, a field-driven optimization control module, a heat flow regulation module, and an energy recovery module.

[0016] (1) TPMS metal support structure module:

[0017] A porous metal support structure composed of triple-periodic minimal curved surfaces Gyroid, Diamond, and Primitive topologies is defined by implicit equations and 3D printed using metal additive manufacturing. The TPMS unit size is 1~30 mm, and the overall porosity can be adjusted within the range of 0.3~0.9 to match different thermal management and energy storage applications.

[0018] (2) Phase change energy storage module:

[0019] The TPMS structure is filled with a phase change material PCM, preferably a paraffin-based, organic composite, or metal-reinforced phase change material. The PCM forms a dense contact interface with the metal skeleton, constituting a TPMS-PCM composite, which is used to achieve latent heat energy storage and enhanced thermal conductivity.

[0020] This module achieves heat absorption and release through latent heat of phase change, and the PCM has an effective thermal conductivity. Determined by both structural porosity and phase transition state:

[0021]

[0022] in, The thermal conductivity of the metal skeleton is... is the matrix thermal conductivity of PCM.

[0023] (3) Temperature field monitoring module:

[0024] A distributed temperature sensor array, preferably a thermistor or miniature thermocouple, is arranged inside the composite and at key interfaces to acquire the temperature distribution inside the composite structure in real time and output a digital temperature field signal, thus establishing a three-dimensional temperature field. This provides foundational data for subsequent field-driven optimization.

[0025] (4) Field-driven optimization control module:

[0026] It is electrically connected to the temperature field monitoring module and is used to optimize the structure and operation based on the temperature field distribution. This module can generate porosity distribution and control instructions based on nonlinear mapping functions and reinforcement learning algorithms, and can export the optimization results as an implicit function file for porosity distribution in the manufacturing stage and design variable porosity actuator drive.

[0027] This module is the core innovative unit of this invention. Based on the real-time temperature field and design objectives, it constructs a nonlinear porosity distribution function:

[0028]

[0029] in, The temperature field driving function can be expressed in power-law form. Alternatively, a deep neural network can be used to achieve complex mapping. When the local temperature rises, the system automatically reduces the porosity to enhance thermal conductivity; in the low-temperature region, it increases the porosity to increase energy storage, achieving optimal coupling between thermal conductivity and energy storage performance.

[0030] (5) Heat flow control module: Located at both ends of the composite, it includes a heating substrate, a heat-conducting base, an insulation layer, and a controllable heat sink, used to realize controlled heat input, heat conduction and cooling output. Its function is to control the direction of heat flow during the charging and discharging process, and realize unidirectional heat conduction and waste heat recovery.

[0031] (6) Energy recovery module: including a thermoelectric power generation unit, used to convert the heat flow difference of the composite into electrical energy, realizing energy recovery and reuse. In some embodiments, the thermoelectric module (Bi2Te3 series) can convert the temperature difference into electrical energy output, realizing the integrated function of thermal storage and power generation.

[0032] As one embodiment of the present invention, the implicit equation includes, but is not limited to, the following forms:

[0033] Gyroid type:

[0034] ;

[0035] Diamond type:

[0036] ;

[0037] Primitive type: ;

[0038] in, The threshold function C(x) is a local function whose value can vary with spatial coordinates to achieve a nonlinear distribution of porosity φ(x). Different porosity distributions can be achieved by adjusting its value. Preferably, there is a single-valued numerical inversion relationship between C(x,y,z) and the target porosity φ, which can be solved through calibration. An implicit model is generated using nTopology or similar design software to achieve customized manufacturing. This TPMS structure has continuous heat conduction channels and a high specific surface area, which can effectively improve the thermal conductivity rate and phase change uniformity.

[0039] In one embodiment of the present invention, the unit size of TPMS is preferably 3 to 15 mm, more preferably 5 mm; the overall porosity φ varies between 0.4 and 0.85, more preferably 0.6 to 0.8, and the structural functional gradient is achieved through nonlinear control.

[0040] As one embodiment of the present invention, (1) the phase change material PCM can be paraffin; (2) the TPMS support skeleton material can be AlSi10Mg; (3) the manufacturing method includes: implicit function derivation → field-driven mapping → slice optimization → 3D printing → post-processing (heat treatment, metal plating, PCM impregnation, vacuum filling and encapsulation).

[0041] As one embodiment of the present invention, vacuum impregnation combined with heat treatment is preferred to reduce PCM leakage and improve contact thermal conductivity.

[0042] In one embodiment of the present invention, the field-driven optimization control module includes: a data acquisition unit (for aggregating distributed temperature, heat flux, and environmental parameters), a temperature field solving unit (capable of solving scalar temperature fields in real time), a nonlinear mapping model (including empirical functions, parameterized power-law models, and exponential mappings f(T)), and a structure control unit (converting the mapping results into an implicit threshold field and outputting an STL). Specifically: the temperature field solving unit supports finite element-based transient simulation for online inference with rapid response; to reduce computational latency, a pre-trained surrogate model can be used to accelerate mapping queries. The structure control unit can output two types of execution paths: one is to export the optimization results as a manufacturing file for the next printing; the other is to drive a field-mounted variable porosity actuator, thereby achieving near real-time porosity adjustment during operation.

[0043] Field-driven optimization principle:

[0044] This invention is based on the principle of temperature field driven design, by establishing a temperature field distribution. With porosity function The nonlinear mapping between them enables bidirectional coupling optimization of the local structure and the thermal field. Under steady-state conditions, the energy conservation equation inside the complex is:

[0045]

[0046] in, For density, For specific heat capacity, The equivalent thermal conductivity is For the latent heat term of phase transition, This represents the liquid phase ratio.

[0047] The temperature field distribution is obtained by solving this equation. This method uses thermal field information as the driving input to generate the optimal porosity distribution. It achieves a direct mapping from thermal field information to geometry, thus significantly reducing the computational cost required for traditional iterative optimization.

[0048] As one embodiment of the present invention, the control algorithm adopted by the field-driven optimization control module is based on deep reinforcement learning, preferably based on the DDPG (Deep Deterministic Policy Gradient) algorithm, wherein: the state space includes at least: local temperature gradient, PCM liquid phase ratio, instantaneous heat flux density, equipment operating parameters and environmental boundary conditions.

[0049] The action space includes at least: local porosity adjustment amount, heat sink operating point, and thermoelectric module power distribution;

[0050] The reward function is a multi-objective weighted sum, including but not limited to: thermal conductivity (melting rate, average heat flux coefficient), energy utilization rate (recoverable heat energy), and temperature uniformity (action smoothness, energy consumption penalty); to accelerate convergence, a model-based prior strategy can be introduced.

[0051] Reinforcement learning optimization mechanism:

[0052] To achieve adaptive control of structural parameters and operational performance, this invention further introduces a deep reinforcement learning mechanism. The control strategy employs the DDPG (Deep Deterministic Policy Gradient) algorithm, which combines environmental conditions and temperature feedback to achieve dynamic optimization.

[0053] State space S: includes the current temperature field distribution, liquid phase ratio, average heat flux density, and boundary conditions;

[0054] Action space A: includes porosity adjustment parameter Δφ and local heat conduction path correction parameter Δk;

[0055] The reward function R, taking into account thermal conductivity η1, energy utilization rate η2, temperature uniformity U, and system stability S0, is defined as follows:

[0056]

[0057] in These are the weighting coefficients.

[0058] By linking with the field drive module, the reinforcement learning controller can correct the structural design parameters in real time during system operation, so that the system can maintain high thermal performance under different operating conditions.

[0059] During system operation, temperature sensors monitor internal temperature changes in real time and feed them back to the control module. The field-driven algorithm dynamically adjusts the local porosity and heat conduction path based on the current temperature field, achieving enhanced local heat conduction and global energy balance.

[0060] In one embodiment of the present invention, the heat flow control module further includes: a controllable heating unit that is electrically heated, used for controllable charging and discharging of the PCM during testing; in another embodiment of the present invention, the energy recovery module is used to convert the temperature difference into an electrical energy recovery device, realizing the storage of waste heat during the day and partial recovery and output at night; the related energy harvesting and inverter system includes an energy acquisition module, a control unit and an inverter circuit to connect weak electrical energy to the grid or supply it to the load; this configuration is particularly suitable for photovoltaic coupling and nighttime power generation scenarios.

[0061] The present invention also provides an optimization method for the aforementioned composite phase change energy storage and thermal management system, comprising the following steps:

[0062] Step 1: Establish a TPMS implicit model and set the initial porosity φ(x) and element parameters (element size, wall thickness, material properties); Step 2: Run steady-state / transient thermal simulations or field temperature measurement experiments under given boundary conditions to obtain the temperature field T(x,t) (measured or predicted by the model); Step 3: Calculate the target porosity distribution based on the temperature field using a nonlinear mapping model f(T).

[0063]

[0064] And perform mechanical / manufacturing feasibility checks (minimum wall thickness, printing resolution, support structure); Step 4, optionally use φ(x) as a supervision signal or initialization policy for a deep reinforcement learning controller to train and obtain the final policy π(·); during training, perform joint optimization using a multi-objective loss function with thermal conductivity, energy storage, and mechanical reliability as optimization objectives (a Pareto front can be obtained using an evolutionary algorithm or multi-objective RL strategy); Step 5, generate manufacturing files based on the final optimization output and perform 3D printing and post-processing, followed by cyclic charging and discharging experiments to verify the improvement of melting time, equivalent thermal conductivity evolution, and total energy storage (experimental verification can be performed using infrared imaging and multi-point temperature sampling to form a comparison dataset).

[0065] As one embodiment of the present invention, the nonlinear evolution of PCM liquid fraction, natural convection promotion effect and equivalent composite thermal conductivity with time and temperature is considered simultaneously during optimization and operation. The liquid phase coupling model is used to correct the temperature field prediction and porosity target mapping.

[0066] During the online operation phase, the nonlinear mapping parameters and RL strategy are dynamically updated using a sliding window based on sensor feedback to cope with environmental disturbances and model uncertainties.

[0067] Experiments show that, under the same heat flux density conditions, the average thermal conductivity of the system of the present invention is improved by about 23% compared with the uniform porous structure, the phase change completion time is shortened by about 25%, and the total energy storage is increased by about 10%. Attached Figure Description

[0068] Figure 1 This is a schematic diagram of the overall system structure of the present invention; Figure 2 This is a schematic diagram of the nonlinear change in porosity driven by a temperature field. Figure 3 This is a structural diagram of the visualization experiment and simulation verification device. Detailed Implementation

[0069] The technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention. The following embodiments are used to illustrate the present invention, but are not intended to limit the scope of the present invention.

[0070] The overall system structure is shown in Figure 1. The triple-period minimal surface structure composite phase change energy storage and thermal management system based on field-driven design of the present invention mainly consists of the following modules:

[0071] TPMS metal support structure module (1): The three-period minimal surface Gyroid, Diamond, Primitive topology generated by implicit equations is printed using selective laser melting (SLM) technology.

[0072] Phase change energy storage module (2): Paraffin phase change material (PCM) is injected into the pores of the metal skeleton to form a continuous heat conduction and energy storage channel.

[0073] Temperature field monitoring module (3): Embedded with multiple thermocouples and infrared array sensors to monitor the internal temperature field of the structure in real time.

[0074] Field-driven optimization control module (4): Receives temperature data and calculates temperature field distribution. And generate porosity distribution based on nonlinear function. .

[0075] Heat flow control module (5): includes heating substrate, heat-conducting base, heat insulation layer and controllable heat sink, used to control the direction of heat charging and discharging.

[0076] Energy recovery module (6): Optionally configured with a thermoelectric power generation unit to convert thermoelectric energy into electrical energy.

[0077] Structural design and parameter setting:

[0078] The three-period minimum surface (TPMS) used in this invention is defined by the following implicit function:

[0079] f(x,y,z) = sin(2πx)cos(2πy)+sin(2πy)cos(2πz)+sin(2πz)cos(2πx) = C(x,y,z)

[0080] in The control parameter ranges from -0.8 to 0.8, corresponding to a porosity variation range of φ = 0.6 to 0.8. By changing the C value, a nonlinear gradient structure from high density (thermal conductivity enhancement zone) to low density (energy storage zone) can be formed.

[0081] The preferred TPMS unit size is 5 mm with a wall thickness of 0.3 mm. The framework material is AlSi10Mg, with a thermal conductivity of 180 W / m·K. The phase change material is paraffin RT58 (melting point 36.8℃, latent heat 180 kJ / kg). In the system design, a field-driven function is used:

[0082]

[0083] in:

[0084]

[0085] α is the control exponent (ranging from 0.6 to 1.2). This function controls the nonlinear change of porosity with temperature, reducing porosity and enhancing thermal conductivity in high-temperature regions, while increasing porosity and improving energy storage in low-temperature regions.

[0086] Simulation optimization and algorithm implementation:

[0087] The field-driven optimization control module adopts a co-simulation framework of Python + nTopology + Ansys, and the calculation process is as follows:

[0088] Input boundary conditions:

[0089] The heat flux density of the heating surface is 3000 W / m², and the fixed temperature of the cooling surface is 293 K.

[0090] Temperature field solution:

[0091] Solve the steady-state heat transfer equation:

[0092]

[0093] in It varies with local porosity.

[0094] Structural update:

[0095] Calculate the temperature of each unit Update porosity:

[0096]

[0097] Convergence condition:

[0098] If ΔT < 0.1 K and Δφ < 0.005, then the design iteration ends.

[0099] Reinforcement learning optimization:

[0100] The DDPG algorithm is introduced to optimize the control parameter α and the boundary porosity range.

[0101] The reward function is:

[0102]

[0103] in To improve thermal conductivity, For energy storage density gain, This refers to temperature nonuniformity.

[0104] Through 200 training iterations, the optimal parameter combination α=0.85, φ was obtained. min =0.42, φ max =0.83, which corresponds to an average thermal conductivity that is about 23% higher than that of a uniform porous structure, a phase change completion time that is about 25% shorter, and an energy storage capacity that is about 10% higher.

[0105] Manufacturing and assembly processes

[0106] Step 1: Structural Forming

[0107] The metal skeleton was printed using SLM technology with a laser power of 200 W, a scanning speed of 800 mm / s, and a layer thickness of 30 μm.

[0108] Step 2, Surface Treatment:

[0109] After printing, heat treatment (500℃×2 h) is performed to eliminate internal stress, and a 5 μm thick Cu coating is applied to improve thermal conductivity.

[0110] Step 3: Phase change material injection:

[0111] The vacuum impregnation method is used to inject molten PCM into the pores of the skeleton under a vacuum of 0.05 MPa, and then seal it after cooling and solidification.

[0112] Step 4: Sensor module embedding:

[0113] The sensors are placed at key nodes inside the structure (center, high-temperature end, low-temperature end) with a sampling frequency of 1 Hz.

[0114] Step 5: System Integration

[0115] The skeleton module is mechanically connected to the heat flow control board, and the control module communicates with the temperature sensor via a data bus.

[0116] Experimental verification and performance analysis

[0117] A charge-dissipation experiment was conducted under a heat flux density of 3000 W / m², and the phase transition temperature range of 45–60℃ was recorded.

[0118] The results show that the field-driven TPMS structure of the present invention significantly improves thermal conductivity and energy storage capacity while maintaining structural stability and cycle life.

[0119] This invention utilizes a field-driven method—temperature field → porosity → structural optimization—to achieve nonlinear distributed design and intelligent control of a TPMS composite phase change energy storage system. The system exhibits excellent performance in terms of enhanced thermal conductivity, improved energy storage efficiency, and energy recovery, demonstrating broad prospects for engineering applications.

[0120] The above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention in any other way. Any modifications or equivalent changes made based on the technical essence of the present invention shall still fall within the scope of protection claimed by the present invention.

Claims

1. A composite phase change energy storage and thermal management system, characterized in that, The system includes the following modules: (1) TPMS metal support structure module: a porous metal support body composed of triple periodic minimal curved surfaces Gyroid, Diamond, and Primitive topologies. The structure is defined by implicit equations and formed by three-dimensional printing using metal additive manufacturing process. The size of the TPMS unit is 1~30 mm, and the overall porosity can be adjusted in the range of 0.3~0.

9. (2) Phase change energy storage module: The TPMS structure is filled with phase change material PCM. The phase change material PCM is paraffin-based, organic composite, or metal-reinforced phase change material. The PCM and the metal skeleton form a dense contact interface to constitute a TPMS-PCM composite. The TPMS metal support structure is defined by implicit equations, which include, but are not limited to, the following forms: Gyroid type: ； Diamond type: ； Primitive type: ; in, It is a local threshold function whose value can vary with spatial coordinates to achieve a nonlinear distribution of porosity φ(x); A single-valued numerical inversion relationship exists between the target porosity φ and the target porosity φ, which can be solved through calibration; and an implicit model can be generated using nTopology or similar design software to achieve customized manufacturing. (3) Temperature field monitoring module: Distributed temperature sensor arrays are arranged inside the composite and at key interfaces. The distributed temperature sensor arrays are thermistors or miniature thermocouples. (4) Field-driven optimization control module: electrically connected to the temperature field monitoring module, used to optimize the structure and operation according to the temperature field distribution; this module can generate porosity distribution and control instructions based on nonlinear mapping function and reinforcement learning algorithm, and can export the optimization results as an implicit function file; The control algorithm used by the field-driven optimization control module is based on the DDPG algorithm, wherein: The state space must include at least: local temperature gradient, PCM liquid phase ratio, instantaneous heat flux density, equipment operating parameters, and environmental boundary conditions; The action space includes at least: local porosity adjustment amount, heat sink operating point, and thermoelectric module power distribution; The reward function is a multi-objective weighted sum, including but not limited to: thermal conductivity, energy utilization, and temperature uniformity; to accelerate convergence, a model-based prior strategy can be introduced. (5) Heat flow control module: disposed at the upper and lower ends of the composite, including heating substrate, heat-conducting base, heat insulation layer and controllable heat sink; (6) Energy recovery module: including thermoelectric power generation unit; The optimization method for the composite phase change energy storage and thermal management system includes the following steps: Step 1: Establish the TPMS implicit model and set the initial porosity φ(x) and element parameters; Step 2: Run steady-state / transient thermal simulations or on-site temperature measurement experiments under given boundary conditions to obtain the temperature field T(x,t); Step 3: Calculate the target porosity distribution based on the temperature field using the nonlinear mapping model f(T). in, The temperature field driving function is expressed in power-law form. Alternatively, a deep neural network can be used to implement complex mappings. When the local temperature rises, the system automatically reduces porosity to enhance thermal conductivity; in the low-temperature region, it increases porosity to increase energy storage, achieving optimal coupling between thermal conductivity and energy storage performance. And perform a mechanical / manufacturing feasibility check; Step 4: Use φ(x) as the supervision signal or initialization policy of the deep reinforcement learning controller to train and obtain the final policy π(·); in the training, perform joint optimization of a multi-objective loss function with thermal conductivity, energy storage and mechanical reliability as optimization objectives. Step 5: Generate manufacturing files based on the final optimized output and perform 3D printing and post-processing. Then, conduct cyclic charging and discharging experiments to verify the improvement of melting time, equivalent thermal conductivity evolution and total energy storage.

2. The composite phase change energy storage and thermal management system as described in claim 1, characterized in that: The TPMS unit size is 3~15 mm, and the overall porosity varies between 0.4 and 0.

85.

3. The composite phase change energy storage and thermal management system according to any one of claims 1 to 2, characterized in that: (1) The phase change material PCM can be paraffin; (2) The TPMS support skeleton material is selected from AlSi10Mg; (3) The manufacturing method includes: implicit function derivation → field-driven mapping → slice optimization → 3D printing → post-processing; vacuum impregnation method combined with heat treatment is used to reduce PCM leakage and improve contact thermal conductivity.

4. The composite phase change energy storage and thermal management system according to claim 1, characterized in that: The field-driven optimization control module includes: a data acquisition unit, a temperature field solving unit, a nonlinear mapping model, and a structural control unit; wherein: the temperature field solving unit supports finite element-based transient simulation for online inference with fast response; to reduce computational latency, a pre-trained surrogate model can be used to accelerate mapping queries; the structural control unit can output two types of execution paths: one is to export the optimization results as manufacturing files for the next printing; the other is to drive the on-site variable porosity actuator, thereby achieving near real-time porosity adjustment during operation.

5. The composite phase change energy storage and thermal management system according to claim 1, characterized in that: The heat flow control module further includes: a controllable heating unit that is electrically heated, used to controllably charge and release heat on the PCM during testing.

6. The composite phase change energy storage and thermal management system according to claim 1, characterized in that: The energy recovery module is used to convert temperature difference into electrical energy recovery, realizing the storage of waste heat during the day and partial recovery and output at night; the related energy collection and inverter system includes an energy collection module, a control unit and an inverter circuit to connect weak electrical energy to the grid or supply it to the load; this configuration is particularly suitable for photovoltaic coupling and nighttime power generation scenarios.

7. The composite phase change energy storage and thermal management system according to claim 6, in the process of optimization and operation, simultaneously considers the nonlinear evolution of PCM liquid phase fraction, natural convection promotion effect and equivalent composite thermal conductivity with time and temperature, and adopts a liquid phase coupling model to correct temperature field prediction and porosity target mapping. During the online operation phase, the nonlinear mapping parameters and RL strategy are dynamically updated using a sliding window based on sensor feedback to cope with environmental disturbances and model uncertainties.

Images