A heat dissipation method and system for servo drives based on artificial intelligence

Abstract

The present invention provides a heat dissipation method and system for a servo drive based on artificial intelligence, which relates to the field of equipment heat dissipation technology. The method includes determining multiple initial heat dissipation installation points based on the infrared working video of the servo drive; determining the heat dissipation effect score of each point based on the infrared working video after the initial installation point is installed; determining multiple sets of subsequent heat dissipation installation point schemes based on the heat dissipation effect score of each point and the infrared working video after the initial installation point is installed; using a generative model to generate multiple sets of subsequent simulated infrared working videos based on the infrared working video of the servo drive, the infrared working video after the initial installation point is installed, and multiple sets of subsequent heat dissipation installation point schemes; determining a target installation point scheme based on the multiple sets of subsequent simulated infrared working videos; installing a heat dissipation device based on the target installation point scheme, and performing heat dissipation control based on the installed heat dissipation device. The method can quickly and accurately determine the target installation scheme of the heat dissipation device.

Description

Technical Field

[0001] The present invention relates to the technical field of equipment heat dissipation, and in particular to a heat dissipation method and system for a servo drive based on artificial intelligence. Background Art

[0002] In modern industrial automation, servo drives are key components for precision motion control and are widely used in applications such as CNC machine tools and robotics. As industry evolves towards higher precision, higher speed, and higher loads, the power density of servo drives is increasing, significantly increasing heat generation during operation. Failure to effectively dissipate heat can lead to excessively high temperatures, degrading the performance and lifespan of internal electronic components, and even causing failures, severely impacting equipment operation and production efficiency. Traditional servo drive cooling methods are often based on empirical design, such as simply increasing the heat sink area or installing fans. This approach fails to accurately analyze actual operating heat distribution and is difficult to adapt to complex operating environments. Insufficient heat dissipation can lead to localized overheating, impacting equipment performance, or excessive heat dissipation, resulting in energy waste and increased costs. While some advanced cooling technologies have been implemented, existing servo drive cooling technologies struggle to comprehensively account for the varying thermal characteristics under varying operating conditions, lacking precise and intelligent planning of heat dissipation installation points and solutions.

[0003] Therefore, how to quickly and accurately determine the target installation plan for heat dissipation equipment is an urgent problem to be solved. Summary of the Invention

[0004] The main technical problem solved by the present invention is how to quickly and accurately determine the target installation solution of the heat dissipation equipment.

[0005] According to a first aspect, the present invention provides a heat dissipation method for a servo drive based on artificial intelligence, comprising: obtaining an infrared working video of the servo drive; determining multiple initial heat dissipation installation points based on the infrared working video of the servo drive; obtaining an infrared working video after the initial installation point is installed; determining a heat dissipation effect score for each point based on the infrared working video after the initial installation point is installed; determining multiple sets of subsequent heat dissipation installation point schemes based on the heat dissipation effect score of each point and the infrared working video after the initial installation point is installed; generating multiple sets of subsequent simulated infrared working videos using a generative model based on the infrared working video of the servo drive, the infrared working video after the initial installation point is installed, and the multiple sets of subsequent heat dissipation installation point schemes; determining a target installation point scheme based on the multiple sets of subsequent simulated infrared working videos; installing a heat dissipation device based on the target installation point scheme, and performing heat dissipation control based on the installed heat dissipation device.

[0006] In one possible implementation, the determining of the target installation point solution based on the multiple sets of subsequent simulated infrared working videos includes: determining multiple excellent heat dissipation points and a set of excellent subsequent heat dissipation installation point solutions based on the multiple sets of subsequent simulated infrared working videos; generating multiple sets of excellent preliminary solutions based on the multiple excellent heat dissipation points and the set of excellent subsequent heat dissipation installation point solutions; constructing a knowledge graph, each knowledge graph including multiple excellent preliminary solution nodes and multiple edges between the multiple nodes, the node features of each excellent preliminary solution node including a set of excellent preliminary solutions and a set of simulated infrared working videos of the excellent preliminary solutions, and the edges between the nodes represent the similarity between the excellent preliminary solutions; processing the knowledge graph based on a graph neural network to determine the target installation point solution.

[0007] In a possible implementation, the target installation point solution is one of multiple sets of excellent preliminary selected solutions.

[0008] In one possible implementation, the generative model is a generative adversarial network.

[0009] According to a second aspect, the present invention provides a heat dissipation system for a servo drive based on artificial intelligence, comprising: a first acquisition module for acquiring an infrared working video of the servo drive;

[0010] an initial heat dissipation point determination module, configured to determine a plurality of initial heat dissipation installation points based on the infrared working video of the servo drive;

[0011] The second acquisition module is used to obtain the infrared working video after the initial installation point is installed;

[0012] A scoring module, configured to determine a heat dissipation effect score for each point based on the infrared working video after the initial installation point is installed;

[0013] A subsequent solution determination module is used to determine multiple sets of subsequent heat dissipation installation point solutions based on the heat dissipation effect score of each point and the infrared working video after the initial installation point is installed;

[0014] a simulated video generation module, configured to generate a plurality of subsequent simulated infrared working videos using a generation model based on the infrared working video of the servo drive, the infrared working video after the initial installation point is installed, and the plurality of subsequent heat dissipation installation point solutions;

[0015] a target solution determination module, configured to determine a target installation point solution based on the multiple sets of subsequent simulated infrared working videos;

[0016] The heat dissipation execution module is used to install the heat dissipation equipment based on the target installation point plan and perform heat dissipation control based on the installed heat dissipation equipment.

[0017] In one possible implementation, the target solution determination module is also used to: determine multiple excellent heat dissipation points and a set of excellent subsequent heat dissipation installation point solutions based on the multiple sets of subsequent simulated infrared working videos; generate multiple sets of excellent preliminary solutions based on the multiple excellent heat dissipation points and the set of excellent subsequent heat dissipation installation point solutions; construct a knowledge graph, each knowledge graph includes multiple excellent preliminary solution nodes and multiple edges between the multiple nodes, the node features of each excellent preliminary solution node include a set of excellent preliminary solutions and a set of simulated infrared working videos of the excellent preliminary solutions, and the edges between the nodes represent the similarity between the excellent preliminary solutions; and process the knowledge graph based on a graph neural network to determine the target installation point solution.

[0018] In a possible implementation, the target installation point solution is one of multiple sets of excellent preliminary selected solutions.

[0019] In one possible implementation, the generative model is a generative adversarial network.

[0020] According to a third aspect, an embodiment of the present invention provides an electronic device, comprising: a processor; a memory; and a computer program; wherein the computer program is stored in the memory and configured to be executed by the processor to implement the method as described above, the method comprising: obtaining an infrared working video of a servo drive; determining multiple initial heat dissipation installation points based on the infrared working video of the servo drive; obtaining an infrared working video after the initial installation point is installed; determining a heat dissipation effect score for each point based on the infrared working video after the initial installation point is installed; determining multiple sets of subsequent heat dissipation installation point schemes based on the heat dissipation effect score of each point and the infrared working video after the initial installation point is installed; generating multiple sets of subsequent simulated infrared working videos using a generative model based on the infrared working video of the servo drive, the infrared working video after the initial installation point is installed, and the multiple sets of subsequent heat dissipation installation point schemes; determining a target installation point scheme based on the multiple sets of subsequent simulated infrared working videos; installing a heat dissipation device based on the target installation point scheme, and performing heat dissipation control based on the installed heat dissipation device.

[0021] According to the fourth aspect, this embodiment provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the aforementioned artificial intelligence-based heat dissipation method for a servo drive, the method comprising: obtaining an infrared working video of the servo drive; determining multiple initial heat dissipation installation points based on the infrared working video of the servo drive; obtaining an infrared working video after the initial installation point is installed; determining a heat dissipation effect score for each point based on the infrared working video after the initial installation point is installed; determining multiple sets of subsequent heat dissipation installation point schemes based on the heat dissipation effect score of each point and the infrared working video after the initial installation point is installed; generating multiple sets of subsequent simulated infrared working videos using a generative model based on the infrared working video of the servo drive, the infrared working video after the initial installation point is installed, and the multiple sets of subsequent heat dissipation installation point schemes; determining a target installation point scheme based on the multiple sets of subsequent simulated infrared working videos; installing a heat dissipation device based on the target installation point scheme, and performing heat dissipation control based on the installed heat dissipation device.

[0022] The present invention provides a heat dissipation method and system for a servo drive based on artificial intelligence, the method comprising: obtaining an infrared working video of the servo drive; determining multiple initial heat dissipation installation points based on the infrared working video of the servo drive; obtaining an infrared working video after the initial installation point is installed; determining a heat dissipation effect score for each point based on the infrared working video after the initial installation point is installed; determining multiple sets of subsequent heat dissipation installation point schemes based on the heat dissipation effect score of each point and the infrared working video after the initial installation point is installed; generating multiple sets of subsequent simulated infrared working videos using a generative model based on the infrared working video of the servo drive, the infrared working video after the initial installation point is installed, and the multiple sets of subsequent heat dissipation installation point schemes; determining a target installation point scheme based on the multiple sets of subsequent simulated infrared working videos; installing a heat dissipation device based on the target installation point scheme, and performing heat dissipation control based on the installed heat dissipation device. The method can quickly and accurately determine the target installation scheme of the heat dissipation device. BRIEF DESCRIPTION OF THE DRAWINGS

[0023] Figure 1 A schematic flow chart of a heat dissipation method for a servo drive based on artificial intelligence provided by an embodiment of the present invention;

[0024] Figure 2 A schematic diagram of a process for determining a target installation point according to an embodiment of the present invention;

[0025] Figure 3 A schematic diagram of a heat dissipation system for a servo drive based on artificial intelligence provided by an embodiment of the present invention; DETAILED DESCRIPTION

[0026] The present invention will be further described in detail below by means of specific embodiments in conjunction with the accompanying drawings. Similar elements in different embodiments are numbered with associated similar elements. In the following embodiments, many detailed descriptions are provided to enable the present invention to be better understood. However, those skilled in the art will readily appreciate that some of the features may be omitted under different circumstances, or may be replaced by other elements, materials, or methods. In some cases, some operations related to the present invention are not shown or described in the specification. This is to avoid the core of the present invention being overwhelmed by excessive descriptions, and for those skilled in the art, it is not necessary to describe these related operations in detail. They can fully understand the related operations based on the description in the specification and the general technical knowledge in the art.

[0027] In an embodiment of the present invention, there is provided Figure 1 A heat dissipation method for a servo drive based on artificial intelligence is shown, and the heat dissipation method for a servo drive based on artificial intelligence includes steps S1 to S8:

[0028] Step S1, obtaining the infrared working video of the servo driver.

[0029] A servo drive is a device that controls a servo motor. It precisely adjusts the motor's position, speed, and torque. In industrial automation, servo drives provide power and motion control support for a variety of devices requiring precise motion control. They are key components for achieving high-precision, high-speed motion in automated production.

[0030] Infrared operating videos are videos captured using infrared imaging technology to show the surface temperature distribution of a servo drive during operation. Infrared operating videos can visually demonstrate the thermal state of a servo drive during operation.

[0031] Step S2: determining a plurality of initial heat dissipation installation points based on the infrared working video of the servo drive.

[0032] In some embodiments, a plurality of initial heat dissipation installation points can be determined based on the infrared working video of the servo drive through an installation point analysis model, wherein the installation point analysis model is a gated loop unit, the input of the installation point analysis model is the infrared working video of the servo drive, and the output of the installation point analysis model is a plurality of initial heat dissipation installation points.

[0033] The Gated Recurrent Unit (GRU) is an optimized recurrent neural network (RNN) structure that addresses the vanishing and exploding gradient problems faced by traditional RNNs when processing sequential data. By introducing update and reset gates to improve the flow and storage of information, the GRU has been widely used in fields such as natural language processing and time series prediction. GRUs are capable of processing long-term dependencies in sequential data and can learn complex temporal patterns in the data.

[0034] The initial heat dissipation installation point is the preliminary installation location point of the heat dissipation device determined by analyzing the infrared working video of the servo drive through the installation point analysis model.

[0035] Heat dissipation devices can be used to reduce the heat generated by devices such as servo drives during operation to ensure that they can operate stably at an appropriate temperature. Heat dissipation device types include heat sinks, cooling fans, etc. Heat dissipation devices can achieve heat dissipation through conduction, convection, radiation, etc.

[0036] The gated recurrent unit has a powerful ability to process time series data. Infrared working videos are image sequences arranged in chronological order, which contain information about the temperature distribution of the servo drive at different times. This information is temporally dependent. The gated recurrent unit can process video data frame by frame through a gating mechanism and effectively capture this temporal dependency. During processing, the gated recurrent unit can remember the temperature change trends of key areas in early video frames and will not lose key information due to the length of the sequence. For example, if the temperature of a certain area continues to rise across multiple frames of video, the gated recurrent unit can identify this trend. By learning the entire video sequence, the gated recurrent unit can analyze the dynamic process of temperature changes to identify areas prone to high temperatures and then determine multiple initial heat dissipation installation points.

[0037] In some embodiments, the installation point analysis model may include a temperature point analysis layer, a heat diffusion path parsing layer, and an installation point strategy generation layer. The input of the temperature point analysis layer is the infrared working video of the servo drive, and the output of the temperature point analysis layer is a high temperature point set, a low temperature point set, and a temperature distribution thermogram sequence. The input of the heat diffusion path parsing layer is a high temperature point set, a low temperature point set, and a temperature distribution thermogram sequence, and the output of the heat diffusion path parsing layer is a heat diffusion critical path diagram, a list of thermal runaway risk points, multiple high temperature area information, and multiple low temperature area information. The input of the installation point strategy generation layer is a heat diffusion critical path diagram, a list of thermal runaway risk points, multiple high temperature area information, and multiple low temperature area information, and the output of the installation point strategy generation layer is multiple initial heat dissipation installation points.

[0038] The high temperature point set is a set of coordinate points whose temperatures exceed a preset threshold value, and represents locations where the surface temperature of the servo drive is higher.

[0039] The low-temperature point set is a set of coordinate points in the infrared working video where the temperature is lower than the preset threshold, reflecting the locations where the servo drive surface temperature is lower and the heat dissipation is better.

[0040] The temperature distribution heatmap sequence is a collection of heatmaps generated frame by frame, visualizing the temperature distribution at various locations on the servo drive surface using color shading. Each image corresponds to a moment in the video.

[0041] The heat diffusion critical path diagram is a schematic diagram formed by analyzing the movement trajectory of high-temperature points and connecting the main paths of heat propagation with arrows. The thickness of the arrows indicates the intensity of heat diffusion.

[0042] The thermal runaway risk point list is a list of coordinate points that screen out the thermal diffusion path, where the points continuously move in the opposite direction toward the low-temperature area and the temperature continues to rise, thus posing an overheating risk.

[0043] The information of multiple high-temperature areas is related information such as boundaries and ranges of multiple high-temperature areas delineated around high-temperature point dense areas and key heat diffusion paths.

[0044] The information of multiple low-temperature areas is related information such as boundaries and ranges of multiple low-temperature areas divided according to the distribution of low-temperature points.

[0045] The temperature point analysis layer mainly extracts high-temperature point sets and low-temperature point sets from the infrared working video of the servo drive, and generates a sequence of temperature distribution thermograms to complete the preliminary analysis and quantification of the temperature information in the video. The heat diffusion path analysis layer can construct a heat diffusion key path diagram to clarify the main channels for heat propagation, and screen out a list of thermal runaway risk points to locate high-risk areas. At the same time, it delineates the boundaries of high-temperature and low-temperature areas, thereby obtaining information on multiple high-temperature areas and multiple low-temperature areas to achieve in-depth analysis of the dynamic diffusion and regional distribution of heat. The installation point strategy generation layer integrates the analysis results of the first two layers, and ultimately determines the initial heat dissipation installation points at heat propagation hubs, risk points, and weak locations within high-temperature areas.

[0046] This layered design breaks down complex problems into three sequential steps: data extraction, path analysis, and solution decision-making. This logically clear approach is easy to understand and optimize. This layered approach also significantly improves model interpretability and scalability, effectively enhancing the efficiency and accuracy of heat sink installation point identification and system stability.

[0047] Step S3, obtaining the infrared working video after the initial installation point is installed.

[0048] The infrared working video after the initial installation point is a video captured using infrared imaging technology after the heat dissipation device is installed on the servo drive according to the initial installation point, reflecting the surface temperature distribution during operation.

[0049] Step S4, determining the heat dissipation effect score of each point based on the infrared working video after the initial installation point is installed.

[0050] In some embodiments, a heat dissipation effect score of each point can be determined using a heat dissipation scoring model based on the infrared working video after the initial installation point is installed. The heat dissipation scoring model is a convolutional neural network model. The input of the heat dissipation scoring model is the infrared working video after the initial installation point is installed, and the output of the heat dissipation scoring model is the heat dissipation effect score of each point.

[0051] Convolutional neural network models include convolutional neural networks (CNNs). Convolutional neural networks (CNNs) are deep learning models that excel at processing grid-structured data, such as images and videos. Convolutional neural networks primarily consist of convolutional layers, pooling layers, and fully connected layers. By leveraging weight sharing and local connectivity, CNNs can effectively reduce model parameters, thereby lowering computational complexity. Convolutional neural networks are widely used in fields such as image recognition and object detection.

[0052] The Thermal Performance Score is a numerical value derived by using a thermal performance scoring model to quantitatively evaluate the heat dissipation capacity of the heat dissipation device at the initial installation point. The Thermal Performance Score converts the abstract concept of thermal performance into a concrete value, facilitating comparison and analysis of heat dissipation at different initial installation points.

[0053] Convolutional neural networks have powerful image feature extraction capabilities, enabling them to determine the cooling effectiveness score for each point based on infrared operating videos taken after initial installation. Infrared operating videos are essentially image sequences, and convolutional neural networks can automatically extract the spatial characteristics of the temperature distribution within the video frames through convolutional layers, such as the shape, size, location, and temperature gradient of hotspots. By combining these features with pooling and fully connected layers, these features can be mapped into a cooling effectiveness score.

[0054] Step S5, determining multiple sets of subsequent heat dissipation installation point solutions based on the heat dissipation effect score of each point and the infrared working video after the initial installation point is installed.

[0055] In some embodiments, a Transformer model can be used to determine multiple sets of subsequent heat dissipation installation point solutions based on the heat dissipation effect score of each point and the infrared working video after the initial installation point is installed. The input of the Transformer model is the heat dissipation effect score of each point and the infrared working video after the initial installation point is installed, and the output of the Transformer model is multiple sets of subsequent heat dissipation installation point solutions.

[0056] The Transformer model is a deep learning model architecture based on the attention mechanism. Using a multi-head attention mechanism, the Transformer model can simultaneously focus on different parts of the input data and capture long-range dependencies, thereby efficiently extracting data features. Widely used in fields such as natural language processing and computer vision, the Transformer model possesses powerful feature representation and modeling capabilities.

[0057] Multiple sets of subsequent heat dissipation installation point solutions are multiple installation solutions for heat dissipation equipment calculated by Transformer to further optimize heat dissipation.

[0058] The heat dissipation effect scores and video data contain complex information. Transformer, through its multi-head attention mechanism, can flexibly weight the key features of this data and discover the intrinsic connection between the scores and the temperature distribution and heat dissipation weak points in the video. This allows it to comprehensively develop multiple sets of subsequent heat dissipation installation point plans to achieve an optimized design of the heat dissipation layout.

[0059] Step S6, generating multiple sets of subsequent simulated infrared working videos using a generative model based on the infrared working video of the servo drive, the infrared working video after the initial installation point is installed, and the multiple sets of subsequent heat dissipation installation point solutions.

[0060] The generative model is a generative adversarial network, the input of which is the infrared working video of the servo drive, the infrared working video after the initial installation point is installed, and the multiple sets of subsequent heat dissipation installation point solutions. The output of the generative model is multiple sets of subsequent simulated infrared working videos.

[0061] A Generative Adversarial Network (GAN) is a deep learning model composed of a generator and a discriminator. The generator creates new data samples, while the discriminator assesses the authenticity of these samples. Through adversarial training, the two continuously improve their performance, ultimately enabling the generator to generate high-quality data.

[0062] The subsequent simulated infrared operation video, generated by the generative model, simulates the servo drive surface temperature distribution after implementing different subsequent heat dissipation installation point solutions. The subsequent simulated infrared operation video can predict the heat dissipation effect after implementing multiple subsequent heat dissipation installation point solutions.

[0063] The generative adversarial network can extract real-world thermal distribution data from infrared operating videos of the servo drive and from videos of the servo drive after initial installation. It can also learn the characteristics and patterns of temperature distribution under different conditions. Multiple subsequent heat dissipation installation point scenarios serve as conditional information, informing the model to generate corresponding results under different heat dissipation strategies. The generator can combine this input information with its learned characteristics and patterns to attempt to generate subsequent simulated infrared operating videos that conform to different heat dissipation installation point scenarios. The discriminator can help the generator continuously adjust, resulting in more realistic and realistic simulated videos.

[0064] Step S7: determining a target installation point plan based on the multiple sets of subsequent simulated infrared working videos.

[0065] In some embodiments, Figure 2 A flowchart of a target installation point determination solution provided by an embodiment of the present invention includes steps S21 to S24:

[0066] Step S21 : determining a plurality of excellent heat dissipation points and a set of excellent subsequent heat dissipation installation point solutions based on the plurality of subsequent simulated infrared working videos.

[0067] In some embodiments, a variational autoencoder can be used to determine multiple excellent heat dissipation points and a set of excellent subsequent heat dissipation installation point solutions based on the multiple sets of subsequent simulated infrared working videos. The input of the variational autoencoder is the multiple sets of subsequent simulated infrared working videos, and the output of the variational autoencoder is multiple excellent heat dissipation points and a set of excellent subsequent heat dissipation installation point solutions.

[0068] A variational autoencoder (VAE) is a generative model that combines autoencoders and variational inference. It consists of an encoder and a decoder. The encoder maps input data to a probability distribution in a latent space. Within this latent space, the VAE assumes the data conforms to a certain probability distribution and samples it using a reparameterization technique. The decoder then converts the sampled latent vector back to the original data. The uniqueness of the VAE lies in its focus on learning the underlying structure and distribution of data while making the latent space continuous and interpretable. Therefore, the VAE can be used for tasks such as data generation and feature extraction.

[0069] The excellent heat dissipation point is a location on the surface of the servo drive where the heat dissipation effect is significantly better than that of other areas, as confirmed by the variational autoencoder.

[0070] A superior heat dissipation installation point solution is selected from multiple heat dissipation installation point solutions after comprehensive consideration of heat dissipation performance. This solution identifies the ideal heat dissipation device installation method and can be subsequently expanded, modified, or combined to identify more valuable heat dissipation installation point solutions.

[0071] The variational autoencoder can learn the latent features of temperature distribution and heat dissipation performance in subsequent simulated infrared operation videos. The encoder maps video data into a latent space and captures key information such as temperature change patterns and hotspot distribution patterns. Because the latent space is continuous, the variational autoencoder can analyze and compare samples in the latent space to identify feature combinations that correspond to excellent heat dissipation performance. The decoder can use these feature combinations to reconstruct the corresponding heat dissipation scenario and identify multiple optimal heat dissipation points. Furthermore, by comprehensively considering the heat dissipation installation point solutions and their effectiveness represented by different samples in the latent space, the variational autoencoder can screen a set of excellent subsequent heat dissipation installation point solutions.

[0072] Step S22 , generating a plurality of sets of excellent preliminary solutions based on the plurality of excellent heat dissipation points and the set of excellent subsequent heat dissipation installation point solutions.

[0073] In some embodiments, a deep neural network model can be used to generate multiple sets of excellent preliminary solutions based on the multiple excellent heat dissipation points and the set of excellent subsequent heat dissipation installation point solutions. The input of the deep neural network model is the multiple excellent heat dissipation points and the set of excellent subsequent heat dissipation installation point solutions, and the output of the deep neural network model is multiple sets of excellent preliminary solutions.

[0074] Deep neural network models are a type of neural network that consists of multiple hidden layers. Through the connections and interactions between numerous neurons, deep neural network models can automatically learn deep features and patterns from complex data. Deep neural network models include deep neural networks (DNNs), which consist of an input layer, multiple hidden layers, and an output layer. DNNs possess powerful nonlinear mapping and generalization capabilities, enabling them to handle complex tasks such as predictive analysis.

[0075] Multiple sets of excellent preliminary solutions are a series of candidate heat dissipation installation point solutions that perform well in many aspects such as heat dissipation effect, cost, and feasibility, generated by a deep neural network model.

[0076] Multiple optimal heat dissipation points and a set of optimal subsequent heat dissipation installation point solutions contain a wealth of heat dissipation-related information. The deep neural network model can deeply analyze and mine this information, learning the inherent relationships and patterns between them. By learning and understanding this information, the model can search and explore the space of potential solutions, generating a variety of different combinations and variations, thereby obtaining multiple sets of preliminary solutions with different characteristics but all performing well. In addition, the generalization ability of deep neural networks enables them to adapt to different practical situations and constraints, making the generated solutions highly versatile and practical.

[0077] Step S23, constructing a knowledge graph, each knowledge graph includes multiple excellent preliminary solution nodes and multiple edges between the multiple nodes, the node features of each excellent preliminary solution node include a set of excellent preliminary solutions and a set of simulated infrared working videos of the excellent preliminary solutions, and the edges between the nodes represent the similarity between the excellent preliminary solutions.

[0078] A knowledge graph is a data structure that graphically displays entities and their relationships. It consists of nodes (vertices) and edges (edges) and can represent complex knowledge in a structured manner. In some embodiments, a knowledge graph can be used to aggregate and present excellent initial solutions and the relationships between them.

[0079] The excellent preliminary solution node is the basic unit in the knowledge graph. Each node represents a specific excellent preliminary solution and contains various feature information related to the solution.

[0080] Node features can be used to describe the properties and characteristics of an excellent preliminary solution node. The node features include a set of excellent preliminary solutions and a simulated infrared working video of the set of excellent preliminary solutions.

[0081] In some embodiments, a simulated infrared working video of each set of excellent preliminary solutions can be generated based on each set of excellent preliminary solutions by generating a model.

[0082] Edges are relationship lines connecting different excellent preliminary solution nodes in the knowledge graph to reflect the degree of association between different excellent preliminary solution nodes. Edges represent the similarity between excellent preliminary solution nodes.

[0083] In some embodiments, the similarity between excellent preliminary selection solutions can be calculated using a deep neural network model.

[0084] Step S24: Process the knowledge graph based on a graph neural network to determine a target installation point solution.

[0085] A graph neural network is a neural network model specifically designed for processing knowledge graph data. It learns and infers the nodes and edges in a graph and aggregates node neighbor information through a message passing mechanism, thereby modeling the structure and properties of the knowledge graph. The input of the graph neural network is the knowledge graph, and the output is a target installation point solution.

[0086] The target installation point solution is a set of installation point solutions that are most suitable for actual application scenarios selected from multiple excellent preliminary solutions by the graph neural network, which comprehensively considers various factors such as heat dissipation effect, installation feasibility, equipment performance, and cost.

[0087] Nodes in the knowledge graph represent excellent preliminary solutions. They carry information about specific solutions and are the fundamental building blocks of the knowledge graph. Node features provide a more comprehensive and comprehensive representation of each solution, allowing the graph neural network to gain a deeper understanding of its characteristics and strengths. Edges connect nodes and visualize the relationships between solutions. Edges use solution similarity as a link to connect dispersed nodes, clearly demonstrating the degree of correlation between solutions. This design allows the graph neural network to discover the commonalities and differences between excellent preliminary solutions.

[0088] Graph neural networks are able to process knowledge graphs to determine target installation point solutions, primarily due to their adaptability and learning capabilities. As a network model designed for knowledge graph structures, graph neural networks can aggregate node neighbor information through a message passing mechanism, allowing them to accurately model excellent preliminary solutions and their similarity relationships within the knowledge graph. Graph neural networks can also automatically mine underlying patterns in the data to extract valuable features, while also possessing powerful nonlinear fitting capabilities, enabling them to capture the complex nonlinear relationships between excellent preliminary solutions. Furthermore, through training and learning, graph neural networks can continuously optimize their parameters and adjust their processing of knowledge graphs, thereby more accurately and reliably determining target installation point solutions.

[0089] Step S8: installing a heat dissipation device based on the target installation point solution, and performing heat dissipation control based on the installed heat dissipation device.

[0090] After the target installation scheme is determined, a heat dissipation device is installed on the servo drive based on the target installation scheme, and heat dissipation control is performed on the servo drive based on the installed heat dissipation device.

[0091] Based on the same inventive concept, Figure 3 A schematic diagram of a heat dissipation system for a servo drive based on artificial intelligence is provided in an embodiment of the present invention. The heat dissipation system for a servo drive based on artificial intelligence includes:

[0092] A first acquisition module 31 is used to acquire infrared working video of the servo driver;

[0093] an initial heat dissipation point determination module 32, configured to determine a plurality of initial heat dissipation installation points based on the infrared working video of the servo drive;

[0094] The second acquisition module 33 is used to obtain the infrared working video after the initial installation point is installed;

[0095] A scoring module 34 is configured to determine a heat dissipation effect score for each point based on the infrared working video after the initial installation point is installed;

[0096] A subsequent solution determination module 35 is configured to determine a plurality of subsequent heat dissipation installation point solutions based on the heat dissipation effect score of each point and the infrared working video after the initial installation point is installed;

[0097] a simulated video generation module 36 for generating a plurality of subsequent simulated infrared working videos using a generation model based on the infrared working video of the servo drive, the infrared working video after the initial installation point is installed, and the plurality of subsequent heat dissipation installation point solutions;

[0098] a target solution determination module 37, configured to determine a target installation point solution based on the multiple sets of subsequent simulated infrared working videos;

[0099] The heat dissipation execution module 38 is configured to install a heat dissipation device based on the target installation point solution and perform heat dissipation control based on the installed heat dissipation device.

[0100] In addition, unless expressly stated in the claims, the order of the processing elements and sequences, the use of alphanumeric characters, or the use of other names described in this specification are not intended to limit the order of the processes and methods of this specification. Although the above disclosure discusses some of the invention embodiments currently considered useful through various examples, it should be understood that such details are for illustrative purposes only, and the appended claims are not limited to the disclosed embodiments. On the contrary, the claims are intended to cover all modifications and equivalent combinations that are consistent with the spirit and scope of the embodiments of this specification. For example, although the system components described above can be implemented by hardware devices, they can also be implemented only by software solutions, such as installing the described system on an existing server or mobile device.

[0101] Similarly, it should be noted that, in order to simplify the presentation of this specification and thus facilitate understanding of one or more embodiments of the invention, the foregoing descriptions of the embodiments of this specification sometimes combine multiple features into a single embodiment, figure, or description thereof. However, this disclosure method does not imply that the subject matter of this specification requires more features than those recited in the claims. In fact, an embodiment may have fewer features than all of the features of a single disclosed embodiment.

[0102] Finally, it should be understood that the embodiments described in this specification are intended only to illustrate the principles of the embodiments of this specification. Other variations may also fall within the scope of this specification. Therefore, by way of example and not limitation, alternative configurations of the embodiments of this specification may be considered consistent with the teachings of this specification. Accordingly, the embodiments of this specification are not limited to the embodiments explicitly described and illustrated in this specification.

Claims

1. A heat dissipation method for a servo drive based on artificial intelligence, characterized in that: include: Get the infrared working video of the servo drive; Determining a plurality of initial heat dissipation installation points based on the infrared working video of the servo drive, wherein determining the plurality of initial heat dissipation installation points based on the infrared working video of the servo drive comprises: determining the plurality of initial heat dissipation installation points based on the infrared working video of the servo drive through an installation point analysis model, wherein the installation point analysis model is a gated cycle unit, and the installation point analysis model comprises a temperature point analysis layer, a heat diffusion path parsing layer, and an installation point strategy generation layer, wherein the input of the temperature point analysis layer is the infrared working video of the servo drive, and the output of the temperature point analysis layer is a high temperature point set, a low temperature point set, and a temperature distribution thermogram sequence, the input of the heat diffusion path parsing layer is a high temperature point set, a low temperature point set, and a temperature distribution thermogram sequence, the output of the heat diffusion path parsing layer is a heat diffusion critical path diagram, a thermal runaway risk point list, a plurality of high temperature area information, and a plurality of low temperature area information, the input of the installation point strategy generation layer is a heat diffusion critical path diagram, a thermal runaway risk point list, a plurality of high temperature area information, and a plurality of low temperature area information, and the output of the installation point strategy generation layer is a plurality of initial heat dissipation installation points; Obtain infrared working video after initial installation of the mounting point; Determine a heat dissipation effect score for each point based on the infrared working video after the initial installation point is installed; Determine multiple sets of subsequent heat dissipation installation point solutions based on the heat dissipation effect score of each point and the infrared working video after the initial installation point is installed; Generate multiple sets of subsequent simulated infrared working videos using a generative model based on the infrared working video of the servo drive, the infrared working video after the initial installation point is installed, and the multiple sets of subsequent heat dissipation installation point solutions; Determining a target installation point solution based on the multiple sets of subsequent simulated infrared working videos, wherein the target installation point solution based on the multiple sets of subsequent simulated infrared working videos includes: Determining a plurality of excellent heat dissipation points and a set of excellent subsequent heat dissipation installation point solutions based on the plurality of subsequent simulated infrared working videos; Generating a plurality of sets of excellent preliminary solutions based on the plurality of excellent heat dissipation points and the set of excellent subsequent heat dissipation installation point solutions; Construct a knowledge graph, where each knowledge graph includes multiple excellent preliminary solution nodes and multiple edges between the nodes. The node features of each excellent preliminary solution node include a set of excellent preliminary solutions and a set of simulated infrared working videos of the excellent preliminary solutions. The edges between the nodes represent the similarity between the excellent preliminary solutions. Processing the knowledge graph based on a graph neural network to determine a target installation point solution; A heat dissipation device is installed based on the target installation point solution, and heat dissipation control is performed based on the installed heat dissipation device.

2. The heat dissipation method of the servo drive based on artificial intelligence according to claim 1, characterized in that: The target installation point solution is one of multiple sets of excellent preliminary solutions.

3. The heat dissipation method of the servo drive based on artificial intelligence according to claim 1, characterized in that: The generative model is a generative adversarial network.

4. A heat dissipation system for a servo drive based on artificial intelligence, characterized in that: include: A first acquisition module is used to acquire infrared working video of the servo driver; an initial heat dissipation point determination module, configured to determine a plurality of initial heat dissipation installation points based on the infrared working video of the servo drive, the initial heat dissipation point determination module further configured to: determine a plurality of initial heat dissipation installation points based on the infrared working video of the servo drive by using an installation point analysis model, the installation point analysis model being a gated loop unit, the installation point analysis model comprising a temperature point analysis layer, a heat diffusion path parsing layer, and an installation point strategy generation layer, the temperature point analysis layer inputting the infrared working video of the servo drive, the temperature point analysis layer outputting a high temperature point set, a low temperature point set, and a temperature distribution thermogram sequence, the heat diffusion path parsing layer inputting a high temperature point set, a low temperature point set, and a temperature distribution thermogram sequence, the heat diffusion path parsing layer outputting a heat diffusion critical path diagram, a thermal runaway risk point list, a plurality of high temperature area information, and a plurality of low temperature area information, the installation point strategy generation layer inputting a heat diffusion critical path diagram, a thermal runaway risk point list, a plurality of high temperature area information, and a plurality of low temperature area information, the installation point strategy generation layer outputting a plurality of initial heat dissipation installation points; The second acquisition module is used to obtain the infrared working video after the initial installation point is installed; A scoring module, configured to determine a heat dissipation effect score for each point based on the infrared working video after the initial installation point is installed; A subsequent solution determination module is used to determine multiple sets of subsequent heat dissipation installation point solutions based on the heat dissipation effect score of each point and the infrared working video after the initial installation point is installed; a simulated video generation module, configured to generate a plurality of subsequent simulated infrared working videos using a generation model based on the infrared working video of the servo drive, the infrared working video after the initial installation point is installed, and the plurality of subsequent heat dissipation installation point solutions; A target solution determination module is configured to determine a target installation point solution based on the multiple sets of subsequent simulated infrared working videos, and the target solution determination module is further configured to: Determining a plurality of excellent heat dissipation points and a set of excellent subsequent heat dissipation installation point solutions based on the plurality of subsequent simulated infrared working videos; Generating a plurality of sets of excellent preliminary solutions based on the plurality of excellent heat dissipation points and the set of excellent subsequent heat dissipation installation point solutions; Construct a knowledge graph, where each knowledge graph includes multiple excellent preliminary solution nodes and multiple edges between the nodes. The node features of each excellent preliminary solution node include a set of excellent preliminary solutions and a set of simulated infrared working videos of the excellent preliminary solutions. The edges between the nodes represent the similarity between the excellent preliminary solutions. Processing the knowledge graph based on a graph neural network to determine a target installation point solution; The heat dissipation execution module is used to install the heat dissipation equipment based on the target installation point plan and perform heat dissipation control based on the installed heat dissipation equipment.

5. The heat dissipation system of the servo drive based on artificial intelligence according to claim 4, characterized in that: The target installation point solution is one of multiple sets of excellent preliminary solutions.

6. The heat dissipation system of the servo drive based on artificial intelligence according to claim 4, characterized in that: The generative model is a generative adversarial network.

Images