Electric robot target detection algorithm and system thereof

Abstract

The invention discloses a power robot target detection algorithm and a system thereof. The method comprises the following steps: S1, a plurality of sensors, a local processor and an edge server are equipped, and a data transmission channel between a power robot and the edge server is provided; s2, generating vivid synthetic data by using a generative adversarial network, and expanding a training data set; s3, loading a pre-trained convolutional neural network model; s4, adaptively adjusting parameters and structures of the deep learning model, and optimizing parameters and strategies of target detection; s5, updating ontology model parameters; s6, realizing an obstacle avoidance function by planning a safe path; and S7, continuously optimizing the reinforcement learning strategy of the deep learning model. According to the method, the vivid synthetic data is generated through the generative adversarial network, the training data set is remarkably expanded, the problems of high acquisition cost and difficulty in labeling of the real data are solved, the synthetic data meeting specific environment characteristics is generated by introducing condition input, and the diversity and representativeness of the data are improved.

Description

Technical Field

[0001] The present invention relates to the technical field of electric robot, and in particular to an electric robot target detection algorithm and a system thereof. Background Art

[0002] With the rapid development of the power industry, power robots are playing an increasingly important role in tasks such as power grid inspection, equipment maintenance, and troubleshooting. However, when performing these tasks, power robots face complex and changing environmental challenges, such as lighting changes, obstructions, and bad weather, which put higher demands on the target detection capabilities of power robots.

[0003] Traditional target detection algorithms mainly rely on a large amount of labeled real data for training, but the real data in the power sector is expensive to obtain and difficult to label, and often has problems such as data imbalance and single features, resulting in limited generalization ability of the trained model and difficulty in adapting to complex and changing environments. In addition, when performing tasks, power robots need to process a large amount of sensor data in real time and perform operations such as target detection, tracking, and obstacle avoidance, which poses a severe challenge to the computing power and real-time performance of power robots.

[0004] In recent years, with the continuous development of deep learning technology, especially the rise of technologies such as generative adversarial networks (GAN) and reinforcement learning, new ideas have been provided for the innovation of target detection algorithms for power robots. GAN can expand the training data set by generating realistic synthetic data to solve the problem of difficulty in obtaining real data; while reinforcement learning can dynamically adjust model parameters and strategies according to environmental feedback to improve the adaptability and robustness of the model.

[0005] However, most of the target detection algorithms for electric robots on the market still have the following problems: First, the data set expansion method is single, and it is difficult to generate synthetic data that conforms to specific environmental characteristics; second, there is a lack of adaptive optimization mechanism, and it is impossible to dynamically adjust model parameters and strategies according to environmental changes; third, the obstacle avoidance function is not accurate enough to ensure the safety and stability of the electric robot during the execution of tasks. Summary of the invention

[0006] The purpose of this section is to summarize some aspects of embodiments of the present invention and briefly introduce some preferred embodiments. Some simplifications or omissions may be made in this section and the specification abstract and the invention title of this application to avoid blurring the purpose of this section, the specification abstract and the invention title, and such simplifications or omissions cannot be used to limit the scope of the present invention.

[0007] In order to solve the above technical problems, the present invention provides the following technical solutions: an electric robot target detection algorithm, comprising the following steps:

[0008] S1: Equipped with multiple sensors, local processors, edge servers, and providing a data transmission channel between the power robot and the edge server;

[0009] S2: Collect environmental data from multiple sensors, synchronize the data in time, and then use generative adversarial networks to generate realistic synthetic data to expand the training dataset;

[0010] S3: Loads a pre-trained convolutional neural network model for preliminary target detection, then runs a preliminary target detection algorithm on the local processor to detect targets in the environment through the convolutional neural network and generate preliminary results;

[0011] S4: Extract environmental features from sensor data, adaptively adjust the parameters and structure of the deep learning model based on the environmental features, and then optimize the parameters and strategies of target detection through reinforcement learning algorithms;

[0012] S5: Assign the preliminary detection results and complex computing tasks to the edge server for processing. The edge server returns the processing results to the local processor and updates the ontology model parameters.

[0013] S6: Use adaptive filtering algorithms to filter environmental noise, use particle filtering technology to achieve real-time tracking of targets, combine sensor data and detection results to identify obstacles in the environment, and achieve obstacle avoidance by planning safe paths;

[0014] S7: Monitor the detection performance of the algorithm in real time and continuously optimize the reinforcement learning strategy of the deep learning model based on the feedback data.

[0015] As a preferred solution of the target detection algorithm of the electric robot described in the present invention, the sensor includes a camera, an infrared sensor and a laser radar, which are used to obtain environmental data.

[0016] As a preferred solution of the target detection algorithm of the electric robot of the present invention, the specific steps of generating the realistic synthetic data are as follows:

[0017] S21: Collect environmental data from various sensors and clean and preprocess the collected data;

[0018] S22: Design a generator and a discriminator network structure. The generator network takes a random noise vector as input to generate realistic synthetic data. The discriminator network inputs real data and generated data and outputs the authenticity probability of the data.

[0019] S23: The generator and discriminator are jointly trained through adversarial training method, the generator and discriminator are optimized respectively using cross entropy loss function, and the model is trained using Adam optimization algorithm;

[0020] S24: Introduce conditional inputs in the generator and discriminator to generate synthetic data that meets specific conditions by combining environmental features with random noise vectors;

[0021] S25: The quality of the generated data was assessed using the structural similarity index and peak signal-to-noise ratio;

[0022] S26: Fuse the generated synthetic data with real data to expand the training data set;

[0023] S27: Use the expanded dataset to train and validate the target detection model, evaluate the performance of the model under different environmental conditions, and adjust the parameters and architecture of the generative adversarial network based on the validation results.

[0024] As a preferred solution of the target detection algorithm of the electric robot described in the present invention, the joint training generator and discriminator includes the following steps:

[0025] S231: Data Preparation

[0026] From the real dataset P data (x) randomly selects a batch of real data

[0027] From the random noise distribution P z (Z) randomly extracts a batch of noise vectors

[0028] S232: Generate data

[0029] Use the generator to generate a batch of synthetic data

[0030] S233: Calculate the discriminator loss

[0031] For each real data x i , calculate the discriminator output D(xi ) ;

[0032] For each generated data Calculate the discriminator output

[0033] Calculate the loss L of the discriminator D ;

[0034] S234: Update the discriminator

[0035] Back propagation calculates the gradient of the discriminator and uses the Adam optimization algorithm to update the parameters of the discriminator;

[0036] S235: Calculate generator loss

[0037] For each generated data Calculate the discriminator output

[0038] Calculate the generator loss L G ;

[0039] S236: Update Generator

[0040] Back propagation calculates the gradient of the generator and uses the Adam optimization algorithm to update the parameters of the generator;

[0041] S237: Repeat steps S231 to S236 until the generator and discriminator achieve the expected performance.

[0042] As a preferred solution of the target detection algorithm of the electric robot described in the present invention, the loss function of the generator is:

[0043]

[0044] in:

[0045] Indicates the number of data in a batch;

[0046] x i represents the i-th real data sample;

[0047] z i represents the i-th noise vector;

[0048] G(z i ) represents the i-th synthetic data generated by the generator;

[0049] D(x i ) represents the discriminator's response to the real data sample x i The output of represents the probability that the sample is judged as real data;

[0050] D(G(z i )) represents the discriminator's response to the generated data G(z i ) output, indicating the probability that the sample is identified as real data;

[0051] logD(x i ) represents the logarithmic probability of true data being identified as true data;

[0052] log(1-D(G(z i ))) represents the logarithmic probability of the generated data being judged as false data;

[0053] The loss function of the generator is:

[0054]

[0055] in:

[0056] m represents the number of data in a batch;

[0057] z i represents the i-th noise vector;

[0058] G(z i ) represents the i-th synthetic data generated by the generator;

[0059] D(G(z i )) represents the discriminator's response to the generated data G(z i ) output, indicating the probability that the sample is identified as real data;

[0060] log(D(G(z i ))) represents the logarithmic probability of the generated data being classified as real data.

[0061] As a preferred solution of the target detection algorithm of the electric robot described in the present invention, the data time synchronization is performed by means of timestamp alignment and data interpolation to ensure that the data of all sensors have temporal and spatial consistency at the same time point.

[0062] As a preferred solution of the electric robot target detection algorithm described in the present invention, the generator and discriminator network structures of the generative adversarial network are multi-layer convolutional neural networks, the generator network includes a deconvolution layer and a batch normalization layer, and the discriminator network includes a convolution layer and a LeakyReLU activation function.

[0063] As a preferred solution of the electric robot target detection algorithm described in the present invention, the method of adaptively adjusting the parameters and structure of the deep learning model includes using a genetic algorithm or a Bayesian optimization method to adaptively select the optimal network structure and parameter combination according to environmental characteristics.

[0064] As a preferred solution of the electric robot target detection algorithm described in the present invention, the detection performance of the real-time monitoring algorithm is achieved by setting multiple performance indicators and dynamically adjusting the training strategy and parameters of the deep learning model according to these indicators.

[0065] A system using the above electric robot target detection algorithm includes:

[0066] A sensor module, wherein the sensor module is used to obtain multi-dimensional data in the environment through a variety of sensors and transmit the data to a local processor;

[0067] A data processing module, which is used to perform time synchronization, cleaning and preprocessing on the collected data to ensure the temporal and spatial consistency of the data;

[0068] A generative adversarial network module, which is used to generate synthetic data through a generative adversarial network and fuse it with real data to expand the training data set and improve the robustness of the model;

[0069] A target detection module, which is used to load a pre-trained convolutional neural network model, perform preliminary target detection on a local processor, and generate a detection result;

[0070] An adaptive adjustment module, which is used to adjust the parameters and structure of the deep learning model according to environmental characteristics and optimize the parameters and strategies of target detection through a reinforcement learning algorithm;

[0071] An edge computing module, which is used to assign preliminary detection results and complex computing tasks to edge servers for processing, and return the processing results to the local processor to update model parameters;

[0072] A real-time tracking module, which tracks the target in real time by using adaptive filtering and particle filtering technology, identifies obstacles in combination with sensor data, and implements obstacle avoidance through path planning;

[0073] The performance monitoring module is used to monitor the detection performance of the algorithm in real time, optimize the reinforcement learning strategy of the deep learning model according to the feedback data, and improve the overall performance of the system.

[0074] Beneficial effects of the present invention:

[0075] 1. This invention generates realistic synthetic data through generative adversarial networks (GANs), which significantly expands the training data set. This method not only solves the problem of high cost of acquiring real data and difficulty in labeling, but also generates synthetic data that conforms to specific environmental characteristics by introducing conditional inputs, thereby improving the diversity and representativeness of the data. At the same time, the quality of the generated data is evaluated using the structural similarity index and peak signal-to-noise ratio, ensuring the high quality of the training data, thereby improving the generalization ability and accuracy of the target detection model.

[0076] 2. The algorithm in the present invention uses a local processor to perform preliminary target detection and assigns complex computing tasks to the edge server for processing. This distributed processing architecture effectively reduces the computing burden of the power robot and improves the real-time performance of target detection. At the same time, the powerful computing power of the edge server can quickly process complex tasks and promptly feed back the processing results to the local processor, realizing efficient target detection and obstacle avoidance functions.

[0077] 3. The algorithm in the present invention can adaptively adjust the parameters and structure of the deep learning model according to the environmental characteristics, and optimize the parameters and strategies of target detection through the reinforcement learning algorithm. This adaptive optimization mechanism enables the electric robot to better adapt to complex and changeable environments, improves the accuracy and robustness of target detection, and at the same time, monitors the detection performance of the algorithm in real time, and continuously optimizes the reinforcement learning strategy of the deep learning model based on feedback data, further improving the overall performance of the algorithm.

[0078] 4. The present invention can extract environmental features from sensor data and adaptively adjust the parameters and structure of the deep learning model. In addition, the parameters and strategies of target detection are optimized through reinforcement learning algorithms, which not only improves the efficiency and accuracy of detection, but also can be dynamically adjusted according to environmental changes to ensure continuous and efficient detection performance.

[0079] 5. The present invention uses adaptive filtering algorithm and particle filtering technology to achieve real-time tracking of the target, and combines sensor data and detection results to identify obstacles in the environment. By planning a safe path, the obstacle avoidance function of the electric robot is realized. This precise real-time tracking and obstacle avoidance capability ensures the safety and stability of the electric robot during the execution of the task. BRIEF DESCRIPTION OF THE DRAWINGS

[0080] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the following briefly introduces the drawings required for describing the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative labor. Among them:

[0081] Figure 1 This is a flowchart of an electric robot target detection algorithm of the present invention. DETAILED DESCRIPTION

[0082] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, the specific implementation methods of the present invention are described in detail below in conjunction with the accompanying drawings.

[0083] In the following description, many specific details are set forth to facilitate a full understanding of the present invention, but the present invention may also be implemented in other ways different from those described herein, and those skilled in the art may make similar generalizations without violating the connotation of the present invention. Therefore, the present invention is not limited to the specific embodiments disclosed below.

[0084] Secondly, the term "one embodiment" or "embodiment" as used herein refers to a specific feature, structure, or characteristic that may be included in at least one implementation of the present invention. The term "in one embodiment" that appears in different places in this specification does not necessarily refer to the same embodiment, nor does it refer to a separate or selective embodiment that is mutually exclusive with other embodiments.

[0085] Secondly, the present invention is described in detail with reference to the schematic diagram. When describing the embodiments of the present invention in detail, for the sake of convenience, the cross-sectional diagrams showing the device structure will not be partially enlarged according to the general scale, and the schematic diagrams are only examples, which should not limit the scope of protection of the present invention. In addition, in actual production, the three-dimensional dimensions of length, width and depth should be included.

[0086] Example 1

[0087] Reference Figure 1 , as a first embodiment of the present invention, provides an electric robot target detection algorithm, comprising the following steps:

[0088] S1: Equipped with multiple sensors, local processors, edge servers, and providing a data transmission channel between the power robot and the edge server;

[0089] The various sensors include cameras, infrared sensors, lidar, etc., which are used to obtain environmental data;

[0090] The local processor is used to perform preliminary data processing and target detection, and the edge server is used to handle complex computing tasks and model updates.

[0091] S2: Collect environmental data from multiple sensors, including images, infrared data, and lidar point cloud data, and synchronize the data in time to ensure the temporal and spatial consistency of the data. Then use the generative adversarial network (GAN) to generate realistic synthetic data, expand the training data set, and improve the robustness of the model.

[0092] Among them, data time synchronization is carried out through timestamp alignment and data interpolation methods to ensure that the data of all sensors have temporal and spatial consistency at the same time point.

[0093] The specific steps for generating the realistic synthetic data are as follows:

[0094] S21: Collect environmental data from multiple sensors (such as cameras, infrared sensors, and lidar), including images, infrared images, and point cloud data, and then clean and preprocess the collected data, such as denoising, normalization, and data enhancement (such as rotation, scaling, and flipping).

[0095] S22: Generative Adversarial Network (GAN) Model Design

[0096] Generator Design: Design a generator network that takes a random noise vector as input and generates realistic synthetic data. For example, the generator can adopt a convolutional neural network (CNN) architecture to map the noise vector to a high-dimensional image space.

[0097] Discriminator design: Design a discriminator network that inputs real data and generated data and outputs the authenticity probability of the data. The discriminator can also use a CNN architecture to distinguish between real data and synthetic data.

[0098] The generator and discriminator network structures of the generative adversarial network are multi-layer convolutional neural networks. Through the multi-layer convolutional neural network structure, high-level feature representations can be extracted and generated layer by layer in the generator and discriminator. This structure helps to deepen the complexity of the network, improve the expressiveness of the model and the authenticity of the generated data.

[0099] The generator network contains deconvolution layers and batch normalization layers. The deconvolution layers in the generator can expand low-dimensional random noise vectors into high-dimensional synthetic data, which helps generate high-resolution, detailed images. The batch normalization layers in the generator can normalize each batch of data, reduce internal covariate shifts, stabilize and accelerate the training process, and improve the quality and diversity of generated data.

[0100] The discriminator network contains convolutional layers and LeakyReLU activation functions. The convolutional layers in the discriminator can extract local features and hierarchical representations of the data, which is very important for identifying the authenticity of the image and detecting the target. The LeakyReLU activation function in the discriminator can avoid the "neuron death" problem of traditional ReLU (that is, some neurons may become completely zero during training and no longer update), while retaining the ability of nonlinear transformation. This can improve the stability and performance of the discriminator and enhance its ability to distinguish between generated data and real data.

[0101] S23: Through the adversarial training method, the generator and the discriminator are jointly trained. The generator continuously generates more realistic synthetic data, while the discriminator continuously improves its ability to distinguish between real data and synthetic data.

[0102] Use the cross entropy loss function to optimize the generator and discriminator respectively. The goal of the generator is to maximize the probability that the discriminator judges that the generated data is real data, while the goal of the discriminator is to minimize the probability of wrong judgment.

[0103] The Adam optimization algorithm is used for model training to ensure the stability and convergence of the training process.

[0104] S24: Introduce conditional inputs into the generator and discriminator, such as environmental features (such as lighting conditions, obstacle distribution, etc.), and generate synthetic data that meets specific conditions by combining environmental features with random noise vectors;

[0105] In the CGAN architecture, the generator and discriminator accept not only random noise or data as input, but also environmental features as additional input to generate more targeted synthetic data.

[0106] S25: Use indicators such as the structural similarity index (SSIM) and peak signal-to-noise ratio (PSNR) to evaluate the quality of the generated data to ensure that the generated data is visually similar to the real data and representative in the feature space;

[0107] Professionals can also be invited to conduct subjective evaluations of the generated data to ensure that the generated data has high credibility under different environmental conditions.

[0108] S26: Fuse the generated synthetic data with real data to expand the training data set. In this way, the robustness of the model in various environmental conditions is improved;

[0109] Before fusing the generated synthetic data with the real data, the generated synthetic data needs to be labeled to ensure that its labels are consistent with the real data. For example, semi-automatic labeling tools can be used to label the generated data to improve labeling efficiency.

[0110] S27: Use the expanded dataset to train and validate the object detection model, evaluate the performance of the model under different environmental conditions, and adjust the parameters and architecture of the generative adversarial network based on the validation results to optimize the quality and diversity of the generated data.

[0111] Through the above steps, a large amount of realistic synthetic data can be generated to expand the training dataset for power robot target detection. These data cover a variety of environmental conditions and improve the robustness of deep learning models.

[0112] S3: Loads a pre-trained convolutional neural network (CNN) model for preliminary target detection, and then runs a preliminary target detection algorithm on the local processor to detect targets in the environment through the convolutional neural network and generate preliminary results.

[0113] S4: Extract environmental features from sensor data and adaptively adjust the parameters and structure of the deep learning model based on the environmental features. For example, in low-light conditions, increase the weight of infrared data. Then, optimize the parameters and strategies of target detection through reinforcement learning algorithms (such as deep Q-network, DQN) to improve detection efficiency and accuracy.

[0114] Among them, the method of adaptively adjusting the parameters and structure of the deep learning model includes using genetic algorithms or Bayesian optimization methods to adaptively select the optimal network structure and parameter combination according to environmental characteristics.

[0115] S5: Assign preliminary detection results and complex computing tasks to edge servers for processing. Edge servers run high-performance computing tasks, such as complex reasoning and model updates of deep models. Edge servers return processing results to local processors and update ontology model parameters.

[0116] S6: Use adaptive filtering algorithms (such as Kalman filtering) to filter environmental noise and improve the accuracy of target detection. Use particle filtering and other technologies to achieve real-time tracking of targets and ensure continuous detection of targets in complex environments. In addition, combine sensor data and detection results to identify obstacles in the environment and achieve obstacle avoidance by planning safe paths.

[0117] S7: Monitor the detection performance of the algorithm in real time, such as accuracy, detection speed, etc., and continuously optimize the reinforcement learning strategy of the deep learning model based on feedback data to improve the overall performance of the system. The detection performance of the real-time monitoring algorithm is monitored by setting multiple performance indicators and dynamically adjusting the training strategy and parameters of the deep learning model based on these indicators.

[0118] Specifically, the joint training of the generator and the discriminator includes the following steps:

[0119] S231: Data Preparation

[0120] From the real dataset P data (x) randomly selects a batch of real data

[0121] From the random noise distribution P z (Z) randomly extracts a batch of noise vectors The noise vector is used as input to the generator to generate new synthetic data;

[0122] S232: Generate data

[0123] Use the generator to generate a batch of synthetic data That is, the generator is used to convert the noise vector into synthetic data, which is used to train the discriminator so that it can distinguish between generated data and real data;

[0124] S233: Calculate the discriminator loss

[0125] For each real data x i , calculate the discriminator output D(x i );

[0126] For each generated data Calculate the discriminator output

[0127] Calculate the loss L of the discriminator D ,The discriminator loss is used to measure the performance of the discriminator in distinguishing real data from generated data. The goal is to maximize the loss so that the discriminator can accurately distinguish between the two types of data;

[0128] S234: Update the discriminator

[0129] Back propagation calculates the gradient of the discriminator and uses the Adam optimization algorithm to update the parameters of the discriminator. Updating the parameters of the discriminator through back propagation and optimization algorithms can improve its ability to distinguish between real data and generated data.

[0130] S235: Calculate generator loss

[0131] For each generated data Calculate the discriminator output

[0132] Calculate the generator loss L G , the generator loss is used to measure the performance of the generator in deceiving the discriminator. The goal is to minimize the loss and make the generated data closer to the real data, thereby deceiving the discriminator;

[0133] S236: Update Generator

[0134] Back propagation calculates the gradient of the generator and uses the Adam optimization algorithm to update the parameters of the generator. By updating the parameters of the generator through back propagation and optimization algorithms, its ability to generate high-quality synthetic data is improved, making it difficult for the discriminator to distinguish between generated data and real data;

[0135] S237: Repeat steps S231 to S236 until the generator and discriminator achieve the expected performance.

[0136] In the process of jointly training the generator and the discriminator, by repeatedly iterating these steps, the generator and the discriminator are continuously optimized, so that the generator can generate more realistic data, and the discriminator can more accurately distinguish between real data and generated data. This adversarial training method ultimately improves the quality of generated data and the robustness of the system.

[0137] As a preferred solution of the target detection algorithm of the electric robot described in the present invention, the loss function of the generator is:

[0138]

[0139] in:

[0140] m represents the number of data in a batch;

[0141] x i represents the i-th real data sample;

[0142] z i represents the i-th noise vector;

[0143] G(z i ) represents the i-th synthetic data generated by the generator;

[0144] D(x i ) represents the discriminator for the real data sample x i The output of represents the probability that the sample is judged as real data;

[0145] D(G(z i )) represents the discriminator's response to the generated data G(z i ) output, indicating the probability that the sample is identified as real data;

[0146] log D(x i ) represents the logarithmic probability that the real data is judged as the real data. The discriminator should output a value close to 1, so the value of this item should be as large as possible.

[0147] log(1-D(G(z i ))) represents the logarithmic probability that the generated data is judged as fake data. The discriminator should output a value close to 0, so the value of this item should be as large as possible.

[0148] The goal of the discriminator is to maximize this loss function to improve its ability to distinguish between real and generated data.

[0149] The loss function of the generator is:

[0150]

[0151] in:

[0152] m represents the number of data in a batch;

[0153] z i represents the i-th noise vector;

[0154] G(z i ) represents the i-th synthetic data generated by the generator;

[0155] D(G(z i )) represents the discriminator's response to the generated data G(z i ) output, indicating the probability that the sample is identified as real data;

[0156] log(D(G(z i))) represents the logarithmic probability that the generated data is judged as real data. The generator hopes that the discriminator will think that the generated data is real, so the value of this item should be as large as possible.

[0157] The goal of the generator is to minimize this loss function to improve the authenticity of its generated data so that the discriminator cannot distinguish between generated data and real data.

[0158] A system using the above electric robot target detection algorithm includes:

[0159] A sensor module, wherein the sensor module is used to obtain multi-dimensional data in the environment through a variety of sensors and transmit the data to a local processor;

[0160] A data processing module, which is used to perform time synchronization, cleaning and preprocessing on the collected data to ensure the temporal and spatial consistency of the data;

[0161] A generative adversarial network module, which is used to generate synthetic data through a generative adversarial network and fuse it with real data to expand the training data set and improve the robustness of the model;

[0162] A target detection module, which is used to load a pre-trained convolutional neural network model, perform preliminary target detection on a local processor, and generate a detection result;

[0163] An adaptive adjustment module, which is used to adjust the parameters and structure of the deep learning model according to environmental characteristics and optimize the parameters and strategies of target detection through a reinforcement learning algorithm;

[0164] An edge computing module, which is used to assign preliminary detection results and complex computing tasks to edge servers for processing, and return the processing results to the local processor to update model parameters;

[0165] A real-time tracking module, which tracks the target in real time by using adaptive filtering and particle filtering technology, identifies obstacles in combination with sensor data, and implements obstacle avoidance through path planning;

[0166] The performance monitoring module is used to monitor the detection performance of the algorithm in real time, optimize the reinforcement learning strategy of the deep learning model according to the feedback data, and improve the overall performance of the system.

[0167] In summary, the present invention generates realistic synthetic data through generative adversarial networks (GANs), significantly expanding the training data set. This method not only solves the problems of high cost of acquiring real data and difficulty in labeling, but also generates synthetic data that conforms to specific environmental characteristics by introducing conditional inputs, thereby improving the diversity and representativeness of the data. At the same time, the quality of the generated data is evaluated using the structural similarity index and peak signal-to-noise ratio to ensure the high quality of the training data, thereby improving the generalization ability and accuracy of the target detection model. The algorithm in the present invention uses a local processor for preliminary target detection and assigns complex computing tasks to edge servers for processing. This distributed processing architecture effectively reduces the computing burden of the power robot and improves the real-time performance of target detection. At the same time, the powerful computing power of the edge server can quickly process complex tasks and promptly feed back the processing results to the local processor, realizing efficient target detection and obstacle avoidance functions. The algorithm in the present invention can adaptively adjust the parameters and structure of the deep learning model according to the environmental characteristics, and optimize the parameters and strategies of target detection through the reinforcement learning algorithm. This adaptive optimization mechanism enables the power robot to better adapt to the complex and changeable environment, improves the accuracy and robustness of target detection, and at the same time, monitors the detection performance of the algorithm in real time, and continuously optimizes the reinforcement learning strategy of the deep learning model according to the feedback data, further improving the overall performance of the algorithm. The present invention can extract environmental features from sensor data, and adaptively adjust the parameters and structure of the deep learning model. In addition, the parameters and strategies of target detection are optimized through the reinforcement learning algorithm, which not only improves the efficiency and accuracy of detection, but also can be dynamically adjusted according to changes in the environment to ensure continuous and efficient detection performance. The present invention uses an adaptive filtering algorithm and particle filtering technology to achieve real-time tracking of the target, and combines sensor data and detection results to identify obstacles in the environment. By planning a safe path, the obstacle avoidance function of the power robot is realized. This precise real-time tracking and obstacle avoidance capability ensures the safety and stability of the power robot in the process of performing tasks.

[0168] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solutions of the present invention may be modified or replaced by equivalents without departing from the spirit and scope of the technical solutions of the present invention, which should all be included in the scope of the claims of the present invention.

Claims

1. A target detection algorithm for an electric robot, characterized in that: The following steps are involved: S1: Equipped with multiple sensors, local processors, edge servers, and providing a data transmission channel between the power robot and the edge server; S2: Collect environmental data from multiple sensors, synchronize the data in time, and then use generative adversarial networks to generate realistic synthetic data to expand the training dataset; S3: Loads a pre-trained convolutional neural network model for preliminary target detection, then runs a preliminary target detection algorithm on the local processor to detect targets in the environment through the convolutional neural network and generate preliminary results; S4: Extract environmental features from sensor data, adaptively adjust the parameters and structure of the deep learning model based on the environmental features, and then optimize the parameters and strategies of target detection through reinforcement learning algorithms; S5: Assign the preliminary detection results and complex computing tasks to the edge server for processing. The edge server returns the processing results to the local processor and updates the ontology model parameters. S6: Use adaptive filtering algorithms to filter environmental noise, use particle filtering technology to achieve real-time tracking of targets, combine sensor data and detection results to identify obstacles in the environment, and achieve obstacle avoidance by planning safe paths; S7: Monitor the detection performance of the algorithm in real time and continuously optimize the reinforcement learning strategy of the deep learning model based on the feedback data.

2. The target detection algorithm for an electric robot according to claim 1, characterized in that: The sensors include cameras, infrared sensors and lidars, which are used to acquire environmental data.

3. The target detection algorithm for electric power robots according to claim 1, characterized in that: The specific steps for generating the realistic synthetic data are as follows: S21: Collect environmental data from various sensors and clean and preprocess the collected data; S22: Design a generator and a discriminator network structure. The generator network takes a random noise vector as input to generate realistic synthetic data. The discriminator network inputs real data and generated data and outputs the authenticity probability of the data. S23: The generator and discriminator are jointly trained through adversarial training method, the generator and discriminator are optimized respectively using cross entropy loss function, and the model is trained using Adam optimization algorithm; S24: Introduce conditional inputs in the generator and discriminator to generate synthetic data that meets specific conditions by combining environmental features with random noise vectors; S25: The quality of the generated data was assessed using the structural similarity index and peak signal-to-noise ratio; S26: Fuse the generated synthetic data with real data to expand the training data set; S27: Use the expanded dataset to train and validate the target detection model, evaluate the performance of the model under different environmental conditions, and adjust the parameters and architecture of the generative adversarial network based on the validation results.

4. The target detection algorithm for an electric robot as claimed in claim 3, characterized in that: The joint training of the generator and the discriminator comprises the following steps: S231: Data Preparation From the real dataset P data (x) randomly selects a batch of real data From the random noise distribution P z (Z) randomly extracts a batch of noise vectors S232: Generate data Use the generator to generate a batch of synthetic data S233: Calculate the discriminator loss For each real data x i , calculate the discriminator output D(x i ); For each generated data Calculate the discriminator output Calculate the loss L of the discriminator D ; S234: Update the discriminator Back propagation calculates the gradient of the discriminator and uses the Adam optimization algorithm to update the parameters of the discriminator; S235: Calculate generator loss For each generated data Calculate the discriminator output Calculate the generator loss L G ; S236: Update Generator Back propagation calculates the gradient of the generator and uses the Adam optimization algorithm to update the parameters of the generator; S237: Repeat steps S231 to S236 until the generator and discriminator achieve the expected performance.

5. The target detection algorithm for an electric robot according to claim 4, characterized in that: The loss function of the discriminator is: in: m represents the number of data in a batch; x i represents the i-th real data sample; z i represents the i-th noise vector; G(z i ) represents the i-th synthetic data generated by the generator; D(x i ) represents the discriminator's response to the real data sample x i The output of represents the probability that the sample is judged as real data; D(G(z i )) represents the discriminator's response to the generated data G(z i ) output, indicating the probability that the sample is identified as real data; logD(x i ) represents the logarithmic probability of true data being identified as true data; log(1-D(G(z i ))) represents the logarithmic probability of the generated data being judged as false data; The loss function of the generator is: in: m represents the number of data in a batch; z i represents the i-th noise vector; G(z i ) represents the i-th synthetic data generated by the generator; D(G(z i )) represents the discriminator's response to the generated data G(z i ) output, indicating the probability that the sample is identified as real data; log(D(G(z i ))) represents the logarithmic probability of the generated data being classified as real data.

6. The target detection algorithm for an electric robot according to claim 1, characterized in that: The data time synchronization is performed by means of timestamp alignment and data interpolation to ensure that the data of all sensors have temporal and spatial consistency at the same time point.

7. The target detection algorithm for an electric robot according to claim 1, characterized in that: The generator and discriminator network structures of the generative adversarial network are multi-layer convolutional neural networks, the generator network includes a deconvolution layer and a batch normalization layer, and the discriminator network includes a convolution layer and a LeakyReLU activation function.

8. The electric robot target detection algorithm according to claim 1, characterized in that: The method for adaptively adjusting the parameters and structure of the deep learning model includes using a genetic algorithm or a Bayesian optimization method to adaptively select the optimal network structure and parameter combination according to environmental characteristics.

9. The target detection algorithm for an electric robot according to claim 1, characterized in that: The detection performance of the real-time monitoring algorithm is achieved by setting multiple performance indicators and dynamically adjusting the training strategy and parameters of the deep learning model based on these indicators.

10. The system of the target detection algorithm of the electric robot according to any one of claims 1 to 9, characterized in that: include: A sensor module, wherein the sensor module is used to obtain multi-dimensional data in the environment through a variety of sensors and transmit the data to a local processor; A data processing module, which is used to perform time synchronization, cleaning and preprocessing on the collected data to ensure the temporal and spatial consistency of the data; A generative adversarial network module, which is used to generate synthetic data through a generative adversarial network and fuse it with real data to expand the training data set and improve the robustness of the model; A target detection module, which is used to load a pre-trained convolutional neural network model, perform preliminary target detection on a local processor, and generate a detection result; An adaptive adjustment module, which is used to adjust the parameters and structure of the deep learning model according to environmental characteristics and optimize the parameters and strategies of target detection through a reinforcement learning algorithm; An edge computing module, which is used to assign preliminary detection results and complex computing tasks to edge servers for processing, and return the processing results to the local processor to update model parameters; A real-time tracking module, which tracks the target in real time by using adaptive filtering and particle filtering technology, identifies obstacles in combination with sensor data, and implements obstacle avoidance through path planning; The performance monitoring module is used to monitor the detection performance of the algorithm in real time, optimize the reinforcement learning strategy of the deep learning model according to the feedback data, and improve the overall performance of the system.

Images