161 results about "Design domain" patented technology

The invention discloses a circuit EMI suppression-oriented multi-knowledge collaborative distillation and optimization method and system, and the method comprises the steps: constructing a basic large language model based on the circuit design of the electronic industry as a student model, and carrying out the adaptive optimization of the circuit design field. Thirdly, integrating the formalized EMI rule base, the simulation data / model base and the expert experience case base, constructing a multi-source heterogeneous teacher knowledge system, and generating a comprehensive guidance signal G; then, candidate circuit design generated by a student model is evaluated through the system, and a multi-objective loss function including basic distillation, rule conformity, performance fitting and expert experience consistency is constructed; according to the method, the universality is reserved by quantifying differences, compliance is ensured through punishment, the performance difference is reduced, dynamic weight adjustment is introduced, and the model performance reaches a preset target through iterative distillation. The model can understand and process terminologies and design rules in the field of circuit design, and the defect that a general AI model lacks domain knowledge is overcome.

The invention provides a topology and morphology integrated optimization design method for a thin-walled structure. The method comprises the steps that 1, a finite element model is established according to design requirements; 2, extracting a thin-wall structure morphology distribution field # imgabs0 # through two-step partial differential equation filtering and Heavide projection operation; 3, constructing a material attribute rational approximation model in combination with the thin-wall morphology distribution field # imgabs1 # and the topological density field v obtained in the step 2, thereby obtaining an elastic modulus E and mass rho in a design domain; 4, performing finite element analysis in combination with the material interpolation model, and calculating the flexibility of the structure; 5, solving to obtain the sensitivity of the structure flexibility and the quality relative to the design variables; step 6, updating the morphology design variable mu and the topology design variable v of the thin plate by adopting a moving asymptote method; and step 7, when a convergence condition is satisfied, stopping optimization iteration and outputting an optimization result. According to the method, the maximum forming angle of the thin-wall structure is controlled by adjusting the filtering radius of the anisotropic filter, and the stamping forming process is met.

The invention belongs to the related technical field of structure optimization, and discloses a full-scale topological optimization method of a structure under dynamic load, which comprises the following steps of: (1) constructing local volume constraints to represent the allowable maximum material volume in each grid neighborhood of the structure to be optimized, and aggregating all the local volume constraints in a design domain into a total constraint; (2) dispersing the duration of the dynamic load, performing dynamic finite element analysis in each time step, taking a continuous density field of the structure as a design variable, minimizing the dynamic flexibility of the structure as a target function, taking a dynamic balance equation of the structure as a control equation, and constructing a constraint function based on global volume constraint and local volume constraint obtained by aggregation, so as to obtain a dynamic load; further constructing a full-scale topological optimization model under the dynamic load; and (3) solving the control equation, carrying out dynamic finite element analysis on the to-be-optimized structure, and solving the full-scale topological optimization model to obtain the optimal density distribution of the structure. The method considers the influence of the dynamic load.

The invention belongs to the field of intelligent manufacturing and engineering design, and particularly relates to an industrial fan structure optimization design method and system based on AI assistance. According to the method, structural components of an industrial fan are modeled into a multi-node topological structure, a dual-channel performance prediction model is constructed, and the mapping relation between the structure and the performance is accurately established. On the basis, a comprehensive objective function is constructed, a boundary penalty mechanism is introduced, a virtual input parameter space is constructed in combination with a structure disturbance sample, a strategy gradient optimization method is adopted to carry out structure search and iterative optimization, and finally a structure scheme meeting boundary conditions and optimal performance is screened out through simulation verification. The system module comprises structure modeling, model training, objective function construction, disturbance generation, optimization execution, simulation verification and result output, and has the advantages of high structure identification precision, high optimization efficiency, strong engineering adaptability and the like.

The invention relates to the technical field of lattice structure topological optimization, and discloses a multi-scale hybrid lattice structure double-variable topological optimization method, which comprises the steps of defining double variables, including a first design variable as a scaling scale and a second design variable as a relative density; determining a basic unit cell of the hybrid lattice structure and obtaining a corresponding unit cell size, wherein the second design variable is determined based on the unit cell size; obtaining a first filling unit cell and a second filling unit cell; determining a design domain, and filling the first filling unit cells and the second filling unit cells into the design domain by taking the flexibility minimization of the hybrid lattice structure as a target and the volume fraction as a constraint; and optimizing the filling process based on a moving asymptote optimization algorithm so as to carry out topological optimization of the hybrid lattice structure based on double variables, and outputting an optimized multi-scale hybrid lattice structure. According to the industrial bearing structure, higher structural specific stiffness, specific strength and energy absorption efficiency are achieved, and the requirements for light weight, self-adaption and high performance in an existing industrial bearing structure can be met.

The invention discloses a thin-wall structure topological optimization design method, device and system, and a storage medium. The method comprises the following steps: calculating any unit density in a design domain according to a structure topological optimization model; calculating a displacement vector according to the density of any unit in the design domain, the total stiffness matrix and the external load vector; calculating the structural flexibility according to the element stiffness matrix and the displacement vector; according to the unit density and the structural flexibility, obtaining the sensitivity of the target function and the constraint function to design variables; establishing a topological optimization model by taking the minimum structural flexibility as a target function, taking the welding spot area as a constraint condition and taking a design variable as a normalized height; and according to the sensitivity of the target function and the constraint function to the design variables and the topological optimization model, performing updating iteration on the design variables by utilizing an MMA algorithm to obtain an optimized structure. By the adoption of the technical scheme, on the premise that the technological constraint of the superplastic forming technology is met, the structure with the excellent mechanical property is obtained.

A method includes defining a design domain for a meta-element configured to be attached to an edge of thin wall structure with an arbitrary boundary condition, and executing a topological optimization process on the design domain and providing a topology optimized shape for the meta-element. The topological optimization process includes an objective function that minimizes a reflection coefficient of flexural waves in the audible frequency range propagating towards and impinging the meta-element.

The invention provides a failure safety topology optimization design method under a dynamic load. The method comprises the following steps: step 1, defining a design domain and design parameters of a structure; 2, performing transient dynamic analysis on the structure; 3, calculating to obtain an equivalent static load of the dynamic load; 4, performing failure safety topological optimization design on the model under the equivalent static load; the maximum flexibility of the minimization structure under the local failure condition is taken as an optimization target; carrying out optimization iteration on the design variable unit density by using a moving asymptote method until the internal loop failure safety topological optimization process is finished; and 5, according to the design variable updated after the failure safety topological optimization, judging whether an outer circulation convergence condition is met or not. According to the method, the possible local failure condition of the structure is fully considered in the design process, and the performance of the structure under the local failure condition is improved; the dynamic load under the original working condition is converted into the static load through the equivalent static load method, and the calculation efficiency of the optimization process is effectively improved.

The present invention discloses a multi-objective thermal-fluid topological design optimization method for the cooling element of a precision gear grinding machine tool, comprising the following steps: Step 1: Define the design domain: Expand the cooling element into a plane, define a two-dimensional design domain representing the cooling element's geometric structure, and discretize it using the finite element method; Step 2: Numerical modeling: Model the cooling element as a porous medium; Under the conditions of laminar incompressible flow, the fluid dynamics are governed by the dimensionless forms of the continuity equation and the momentum equation; The fluid-solid coupled heat transfer model is governed by the dimensionless energy conservation equation; Step 3: Construct an objective function: The optimization objective is defined as a weighted function of heat transfer and fluid dissipated power, and a fluid-solid heat transfer topological optimization model is constructed; Step 4: Solve the objective function; Step 5: Restore the shape and structure of the cooling element. The present invention also discloses a precision gear grinding machine tool and a cooling element.

This invention discloses an optimized design method for support structures to control thermal stress in metal additive manufacturing, relating to the field of metal additive manufacturing. The method includes the following steps: defining a design domain and a non-design domain; initializing design variables; converting design variables into physical variables through density filtering and projection; establishing an interpolation model between the physical variables and the overall structural stiffness; performing finite element simulation on the layer-by-layer additive manufacturing process of the overall structure to obtain the thermal deformation of the overall structure; and calculating the maximum thermal stress σ of the structure. pn The proposed method involves calculating the volume fraction V of the supporting structure and the suspension angle constraint response G; determining the sensitivity; updating the design variables to solve the optimization model; determining whether convergence has occurred; and post-processing the optimization results. The structure designed by this method ensures that the thermal stress of the overall structure is controllable after additive manufacturing, avoiding cracking and delamination caused by excessive thermal stress.

The invention relates to the field of computer-aided engineering and intelligent manufacturing, and provides a topological optimization design method integrating subjective preferences. The method comprises the following steps: initializing material density distribution in a design domain, and intelligently recommending an initial density distribution scheme according to a subjective preference type selected by a user; calculating an objective response value and sensitivity thereof; calling a pre-trained subjective preference scoring model to subjectively evaluate the current material density distribution, outputting a subjective evaluation value, and calculating the sensitivity of the subjective evaluation value based on a numerical perturbation method; the objective response value, the sensitivity of the objective response value, the subjective evaluation value and the sensitivity of the subjective evaluation value are input into an MMA optimizer for iterative calculation, and material density distribution is dynamically updated to balance subjective preference and objective constraint; and judging whether the updated material density distribution meets a convergence condition set by a user, if so, performing post-processing visualization on the optimized material density distribution, and outputting a final topological configuration, otherwise, returning to a new round of iteration.

The invention relates to the technical field of lattice structure optimization, and discloses a topological optimization-based lattice structure, a design method and application, a topological optimization model is constructed, and physical parameters and optimization parameters of the topological optimization model are initialized; defining a structural design domain, a load and a boundary condition corresponding to the topological optimization model, and establishing a finite element model; the defined load and boundary conditions are applied to the topological optimization model, finite element analysis is carried out, displacement response of each node on a design domain is obtained, and sensitivity of structural flexibility and volume about design variables is analyzed; filtering the sensitivity, and filtering the material density represented by the design variable by adopting a projection method based on a Heaviside function so as to inhibit a checkerboard phenomenon; updating a design variable according to the filtered sensitivity and material density; and checking whether the current iteration result meets a convergence condition, if so, outputting the result, and if not, returning to continue iteration.

The invention belongs to the field of structure optimization design, and particularly discloses a multi-model topological optimization method and system for a complex geometric model, and the method comprises the steps: building a corresponding finite element model according to a geometric model of a wing corrugated structure; dividing a design domain for the geometric model, and establishing a topological model corresponding to the finite element model; geometric mapping is carried out on the shared design domain through manufacturing constraints with repeated modes, a mapping relation of a local coordinate system for processing a large deformation area is established, a topology model capable of being synchronously optimized is formed, and an output file of the topology model is exported; and carrying out synchronous loop optimization on the plurality of topology models by utilizing a parallel computing interface, and judging optimization results of the topology models according to a convergence criterion until the topology models reach the convergence criterion, so as to obtain a final optimization model. According to the method, the geometric mapping effect and the multi-working-condition synchronous optimization effect of the complex geometric model can be improved.

The invention relates to the field of model design, and provides an aircraft multi-objective optimization design method and system based on semanteme in order to improve the calculation efficiency of multi-objective balance. Firstly, original design parameters and design rules are obtained, obtained data are converted into semantic vectors through a semantic embedding function, and adjacency matrixes are obtained; generating a semantic gene map based on the semantic vector and the adjacent matrix, constructing a comprehensive objective function, and performing optimization solution based on the semantic gene map; and creating a target mapping function, and mapping an optimized result into a specific engineering target based on the target mapping function. By introducing a semantic modeling and cognitive enhancement mechanism, aircraft structure optimization is improved from traditional numerical drive to semantic drive, a prediction and search process is accelerated by using semantic space and cognitive reasoning, and global optimization and traceability are realized in combination with crowd intelligence evolution and a cross-domain hypernetwork; not only are efficiency and robustness improved, but also engineering feasibility and compliance credibility of a design result are ensured.

The invention discloses a method for optimizing manufacturability of a welding topological optimization structure, which belongs to the technical field of computer aided design and comprises the following steps: acquiring boundary conditions of a design domain and a geometric model in the design domain; performing initial layout on the geometric model in the design domain through the quadrilateral component; the overlapped quadrilateral assemblies are operated, and segmented and recombined convex polygon assemblies are obtained; acquiring a position relationship between the network nodes in the design domain and the convex polygon component, and converting the position relationship into a unit density field; establishing a topological optimization model according to the unit density field; updating control parameters of the quadrilateral component according to the sensitivity of the target function and the sensitivity of the constraint function; and outputting the layout of the convex polygon component in the design domain after the control parameters are optimized as an optimization result until the topological optimization model converges. According to the optimization method for the manufacturability of the welding topological optimization structure, the problems that existing welding topological optimization is poor in manufacturability and complex in structure post-processing are solved.

The invention provides a neural network multi-objective optimization and FPGA hardware acceleration collaborative design method, and belongs to the field of deep learning model compression and hardware collaborative design. The method comprises the following steps: constructing a joint optimization space containing a neural network compression parameter and an FPGA hardware design parameter; a multi-target Bayesian optimization search strategy is adopted, iterative search is carried out in the joint optimization space, model precision, FPGA resource occupation and reasoning delay are synchronously optimized, and optimal candidate configuration is obtained; matching the compressed network structure with the FPGA parallel architecture by using a hardware-perceived pruning and quantification strategy; a multi-task performance prediction model is adopted to quickly predict the precision, resource occupation and delay of the optimal candidate configuration so as to accelerate the search process; according to the optimal configuration, a hardware accelerator code facing the target FPGA is automatically generated, and integration and implementation are completed. According to the method, collaborative optimization of neural network compression and hardware design is achieved, FPGA resource occupation can be remarkably reduced, the reasoning speed can be increased, and meanwhile the model precision is kept.

The invention discloses a structural topology optimization design method for additive manufacturing and a related device, which can be used in the field of structural design, in the method, based on the shape of a structural design domain and a parameterized density field function including the structural design domain, a structural density field is determined, and a suspension angle constraint function at a structural boundary is constructed; constructing a boundary gradient constraint function based on a normal vector pointing to the interior of the structural design domain on the boundary of the structural design domain; discretizing the structural design domain into a plurality of finite element units, and constructing a material stiffness matrix to calculate an overall stiffness matrix; and constructing a topological optimization mathematical model according to the finite element balance constraint, the volume constraint, the suspension angle constraint function and the boundary gradient constraint function based on the overall stiffness matrix, and solving to obtain a structural topological optimization result. Therefore, through multi-constraint collaborative optimization, the conflict problem of suspended constraint and material anisotropy on the forming direction is solved, and a structural topological optimization result considering manufacturability and structural performance is obtained.

The invention discloses a direct topology and shape optimization method for an existing engineering structure model, and relates to the technical field of topology optimization and shape optimization, and the method comprises the steps: carrying out the mapping through a regular hexahedron unit based on a finite element simulation model of an original engineering structure, and determining an approximate optimization simulation model; based on the approximate optimization simulation model, determining a typical design mode of multi-component non-interference optimization by transforming the state and range of a design domain; based on the typical design mode, the same discrete adjoint equation is adopted to convert the sensitivity of independent or combined application under different sensitivity algorithms, and a compatible unified optimization framework is determined; and determining a variable grouping strategy based on the compatible unified optimization framework so as to uniformly apply various manufacturing process constraints, and obtaining a structure improvement scheme with high manufacturability. The typical design mode is realized by reconstructing the original optimization simulation model, the unified discrete adjoint equation is constructed to realize free combination of different optimization algorithms, and the optimization efficiency and the engineering practicability of multi-component design are improved.

The invention relates to the field of scroll compressor design, and discloses a full-meshing double-vortex-ring scroll compressor diffusion hole machining design method which comprises the steps of constructing an initial design domain, defining a multi-working-condition working point set and a fluid dynamic control equation determined by compressor operation parameters to construct a multi-working-condition analysis model, establishing a solid isotropic material penalty model, and designing a diffusion hole of a full-meshing double-vortex-ring scroll compressor according to the solid isotropic material penalty model. Multi-working-condition robustness topological optimization with the total pressure loss weighted sum under all the weighted working conditions as the target is executed, a discrete pseudo-density field is obtained, parameterized geometric reconstruction is conducted on the discrete pseudo-density field, and a diffuser hole parameterized geometric model controlled by parameter vectors is obtained. By introducing multi-working-condition robustness topological optimization and integrated dynamic machining constraint linkage re-optimization, remarkable robustness of aerodynamic performance of the diffusion hole in a wide working range is achieved, it is ensured that the diffusion hole meets the high-quality and high-efficiency numerical control machining requirements congenitally, and the design efficiency and manufacturing feasibility of the diffusion hole of the scroll compressor are greatly improved.

The invention discloses a multi-scale structural stiffness optimization design method based on neural network re-parameterization, which comprises the following steps: microstructure parameterization modeling: carrying out parameterization modeling on a microstructure by adopting a level set method, and controlling the volume fraction and geometric configuration of the microstructure by adjusting a cutting height parameter; proxy model construction: establishing a mapping relation between cutting height parameters and microstructure equivalent mechanical properties by using a neural network to form a high-precision proxy model; neural network re-parameterization design: constructing a neural network optimizer, taking coordinates of each discrete unit in a design domain as input, and outputting corresponding cutting height parameters; neural network parameters are updated through automatic differentiation and back propagation, and continuous optimization of design variables is achieved; and a multi-scale topological optimization process: calculating a global stiffness matrix and structural flexibility in combination with an agent model and finite element analysis, and obtaining an optimal multi-scale structural topology meeting volume constraints through loss function minimization iterative optimization.

According to the gluing track design method, the storage medium and the electronic equipment, the to-be-glued area is divided into the multiple period design domains, after the gluing starting position is selected, the multiple gluing tracks in each period design domain are planned according to the genetic optimization algorithm, and according to the fact that the coverage rate corresponding to the gluing result is not smaller than the target gluing coverage rate, the gluing track is designed. Under the conditions that the gluing thickness corresponding to the simulated gluing result is consistent with the target gluing thickness and the gluing amount corresponding to the simulated gluing result does not exceed the target gluing amount, the optimal gluing track is selected from the multiple gluing tracks to serve as the final gluing track. According to the scheme, based on a genetic optimization algorithm framework, a gluing track parameterization expression method of a multi-cycle design domain is established, gene crossover and gene variation mechanisms are improved, and a gluing track planning method with the coverage rate, the gluing thickness and the gluing amount meeting target requirements is provided. According to the scheme, the gluing track meeting the requirements for the coverage rate and the glue amount can be obtained, and the advantages of being high in coverage rate and low in glue amount are achieved.

The embodiment of the invention provides a method and equipment for automatically designing a domain model, and a storage medium. The method comprises the following steps: acquiring a plurality of business object names input by a user; constructing a first prompt instruction used for prompting design of the domain models, and inputting the first prompt instruction and the plurality of business object names into the first large language model to obtain JSON description of the domain models corresponding to the business object names and JSON description of a data flow relationship between the domain models; and constructing a second prompt instruction used for prompting design of a model field of the domain model, and inputting the second prompt instruction, the JSON description of the domain model and the JSON description of the data flow relationship related to the domain model into a second large language model to obtain the JSON description of the model field of the domain model. The domain model design is automatically generated through a large language model, a user only needs to input a business object name, a system can automatically output the domain model in a JSON format, a model data flow relation and a model field definition, and the domain model and related description thereof are efficiently output.

This invention belongs to the field of artificial intelligence hardware architecture design and discloses a method, system, device, and medium for cross-layer optimization of multi-core packaging and optical interconnect clusters for large language model training. The method includes: constructing a design space, including a core hardware architecture layer, an optical interconnect network layer, and a training parallel strategy layer; performing an outer-layer search to search for the architecture parameters of the multi-core modules within the core hardware architecture layer; for each outer-layer search sampling point, performing an inner-layer search to collaboratively optimize the optical interconnect network topology and training parallel strategy within the optical interconnect network layer and the training parallel strategy layer; and outputting the Pareto optimal design point on the training performance and cluster cost plane. The technical solution described in this invention can guide the design of related training clusters, thereby fully leveraging the advantages and potential of core technology and optical interconnect technology in large language model training.

The application provides a three-period minimal surface intervertebral fusion cage design method, relates to the technical field of porous structure design, and determines a design domain of a three-period minimal surface TPMS porous structure region, combines individual vectors of modeling parameters of corresponding TPMS of each region, randomly generates a plurality of individual vectors, and forms an initial population; finite element simulation is carried out on the TPMS porous model corresponding to each individual vector, so as to divide the initial population into a load advantage population and a pore advantage population; the individual vectors are selected from the load advantage population and the pore advantage population respectively to perform improved neighborhood directional crossover operation; an optimal three-period minimal surface TPMS porous structure region is obtained through iterative genetic algorithm; through improved crossover operation and neighborhood directional crossover between the load advantage population and the pore advantage population, the new generation of individuals can consider both load performance and pore performance, and optimization is avoided around a single performance.

The invention discloses a topological optimization method and device considering volume constraints, boundary conditions and a design domain based on deep learning, and belongs to the technical field of topological optimization design. According to the method, training data sets under various constraint conditions and the design domain are constructed, a full convolutional neural network is adopted for offline training, and the training data sets are optimized; and obtaining a full convolutional neural network model capable of representing a topological optimization design rule. In practical application, only corresponding boundary conditions and design domain information need to be input, a topological optimization result can be rapidly output through the trained neural network, and therefore a traditional iterative calculation process is replaced. According to the method, the calculation complexity and the time consumption are remarkably reduced while the topological optimization precision is kept, high efficiency and intelligence of topological optimization are realized, and the method has a good engineering application prospect.

The present invention relates to a topology optimization method for a functionally graded structure based on elastic hysteresis, comprising: defining the hysteresis energy loss; defining the design variables; discretizing the design domain into a number of regions and defining the density of the regions as the local density; taking the relative density of the overall structure as the design variable, taking the minimum structural compliance as the objective, adding corresponding constraints to the local density of each region, and at the same time taking the hysteresis energy loss and volume as constraints to establish a multi-level functionally graded structure topology optimization model; iteratively solving the multi-level functionally graded structure topology optimization model to obtain the optimal design variables that meet the optimization objective, and obtaining the optimal topological structure based on the optimal design variables. The purpose of the method of the present invention is to reduce the hysteresis loss of the structure while ensuring the continuity of the structural units after optimization.

A 4D printing intelligent structure design method based on a convolutional neural network comprises the following steps: parametric modeling and data generation: controlling finite element software through a script to construct a'material parameter field-target deformation field 'sample data set in batches; data preprocessing and geometric standardization: uniformly converting a complex design domain into a standard rectangular domain by adopting a conformal mapping technology, and realizing high-fidelity standardization of physical field data; constructing a U-Net + ResNet hybrid convolutional neural network, and combining with composite loss function training to synchronously optimize global and key region deformation prediction precision; performing standardization processing on the new design after model deployment to realize second-level high-precision deformation prediction; the physical space design is restored through inverse mapping, multi-material 4D printing manufacturing is driven, and a deformation verification closed loop is completed. According to the method, the limitation of complex geometric boundaries is broken through, the nonlinear and multi-physics field coupling characteristics of the intelligent material can be accurately represented, and the practicability of the scheme is ensured by the constructed'design-prediction-manufacturing 'integrated process.

The invention discloses a topological optimization method and device based on a CutFEM and an MMC, a medium and equipment. The method comprises the steps that a design domain is defined, a plurality of initial moving deformation assemblies are arranged in the design domain, and material parameters of the moving deformation assemblies are set; carrying out structured background grid unit division on the design domain; dividing each background grid unit into an entity unit, a blank unit and a cutting unit, and performing sub-grid division on a material area of the cutting unit; identifying boundary lines between the material areas and the material-free areas of the cutting units as inner boundaries and outer boundaries; carrying out stiffness matrix assembly on the material areas of the entity unit and the cutting unit, forming and solving a total stiffness matrix equation, and obtaining node displacement; generating and solving a topological optimization column formula; and calculating the sensitivity, and carrying out iterative optimization. The method can solve the problems that an existing MMC method is low in efficiency and limited in precision under the large-scale complex geometric boundary condition.

The present invention provides a method for failure-safe topology optimization design under dynamic load, comprising: step 1, defining the design domain and design parameters of the structure; step 2, performing transient dynamic analysis on the structure; step 3, calculating the equivalent static load of the dynamic load; step 4: performing failure-safe topology optimization design on the model under equivalent static load; minimizing the maximum flexibility of the structure under local failure as the optimization goal; and using the moving asymptote method to iteratively optimize the design variable unit density until the inner loop failure-safe topology optimization process is completed; step 5: judging whether the outer loop convergence condition is met based on the updated design variables after the failure-safe topology optimization. The present invention fully considers the possible local failure of the structure during the design process, thereby improving the performance of the structure under local failure conditions; converting the dynamic load under the original working condition into a static load through the equivalent static load method, effectively improving the computational efficiency of the optimization process.

The application discloses a kind of based on NiTi alloy's chip liquid cooling plate flow channel topological optimization design method, comprising: obtaining chip heat source distribution, determine two-dimensional design domain and establish flow-thermal coupling numerical model;Determine material, environment and boundary parameter;Based on solid isotropic material interpolation format, express physical property parameter as the function of design variable, construct topological optimization mathematical model;With minimum fluid viscosity energy dissipation rate and design domain average temperature as joint target, with fluid volume fraction and temperature interval as constraint, carry out iterative solution, obtain two-dimensional optimal flow channel topological configuration.Wherein, NiTi shape memory alloy unit is arranged on the wall surface of flow channel, adaptive flow regulation based on temperature sensing can be realized.The application realizes the collaborative optimization of heat dissipation performance and flow resistance, can effectively suppress hot spot, and design efficiency is high.