Bionic robotic fish swimming control method based on superficial brain model

Abstract

The invention relates to a bionic robotic fish swimming control method based on a superficial brain model, and relates to the technical field of robotic fish, and the method comprises the steps: building a superficial brain model through simulating a biological nervous system and combining a biological neuron information transmission mechanism, and improving the motion ability of the bionic robotic fish; the superficial brain model is mainly used for building a hippocampus body module and a cerebellum module to simulate a biological nervous system of a real organism, so that the bionic robotic fish is more like a real fish. By adopting a recursive fractal structure, the efficiency of neural computation is improved, and the robustness of the network is enhanced through a randomized connection mode.

Description

Technical Field

[0001] The present invention belongs to the technical field of bionic fish, and particularly relates to a method for controlling the swimming of a bionic robotic fish based on a shallow brain model. Background Technique

[0002] The statements in this part only provide background technical information related to the present invention and do not necessarily constitute prior art.

[0003] The ocean covers about 70% of the earth's area and has always contained rich resources. The underwater world is full of unknown opportunities and challenges. In recent years, bionic robots have gradually become a hot field in scientific research. As one of the representative applications, underwater bionic robotic fish have broad application prospects in the fields of ocean exploration, environmental monitoring, military security, etc. By simulating the behavior of real fish, it provides a new idea for us to deeply understand the motion control of organisms.

[0004] Traditional bionic robotic fish usually rely on complex mechanical structures and control algorithms to achieve basic swimming functions. Although there have been great breakthroughs in the structural design and motion ability of bionic robotic fish at present, how to achieve efficient, accurate and low-energy-consuming swimming control is still an important problem that needs to be continuously solved. Many existing control algorithms are often based on preset rules, lacking sufficient flexibility and adaptability, and unable to respond quickly to environmental changes like organisms in nature. Secondly, traditional robotic fish use multiple motors to drive multiple joints, with frequent starts and stops, and the algorithms are not optimized enough, resulting in high energy consumption. The introduction of the brain-like model has opened a new chapter in the control of bionic robotic fish. By mimicking the processing method of the brain neural network, the brain-like model can achieve efficient control algorithms, thus promoting the development of bionic robotic fish towards the direction of intelligence and high efficiency. Summary of the Invention

[0005] The present invention starts from the reasons for the swimming control scheme of the robotic fish:

[0006] To solve the above problems, the present invention provides a method for controlling the swimming of a bionic robotic fish based on a shallow brain model. The present invention constructs a shallow brain neural network model. By simulating the biological nervous system and combining the information transmission of biological neurons for swimming control, the perception and motion ability of the robotic fish in a dynamic environment are improved. Obstacles are randomly arranged in the simulation environment. The bionic robotic fish receives sensor data through the perception input layer. The perception input layer includes at least a distance sensor and an angle sensor to obtain the relative distance and angle information between the robotic fish and the surrounding environmental obstacles. The data is transmitted to the neural computing layer and processed through the hippocampus module and the cerebellum module. The output control layer converts the motion control signal generated by the neural computing layer into the motion instruction of the robotic fish to ensure that the robotic fish avoids obstacles and maintains stable swimming.

[0007] To achieve the above object, the present invention adopts the following technical solutions:

[0008] A bionic fish swimming control method based on a shallow brain model, comprising the following steps:

[0009] Construct a shallow brain model, including a perception input layer, a neural computation layer, and an output control layer. The neural computation layer mainly includes a hippocampus module and a cerebellum module;

[0010] The perception input layer is used to receive the input data of the sensors. The sensors include at least one distance sensor and one angle sensor. The input data includes the relative distance information and angle information between the bionic fish and the surrounding environment;

[0011] The neural computation layer mimics the basic working mechanisms of certain specific regions of the brain. According to the brain functional partition, especially for performing specific tasks such as perception, decision-making, and motion control, a hippocampus module and a cerebellum module are constructed to perceive the input data and generate output signals and motion control signals;

[0012] Simulate the neuron connection mode of real organisms. Neurons between each layer are randomly connected, and neurons within the layer do not interfere with each other.

[0013] The output control layer generates the motion instructions of the bionic fish according to the control signals generated by the neural computation layer, and drives the caudal fin and pectoral fins of the bionic fish to perform swimming actions, realizing swimming control and obstacle avoidance.

[0014] The shallow brain model is used for simulating the bionic fish swimming control to obtain the motion trajectory of the bionic fish.

[0015] As an alternative implementation, use matlab for simulation, encapsulate the hippocampus and cerebellum models, and build the hierarchical connection of neurons. Generate obstacles in the coordinates and distribute them randomly.

[0016] As an alternative implementation, the neural computation layer of the shallow brain model adopts a recursive fractal neural network structure to enhance the computational efficiency of the neural network through hierarchical network design.

[0017] As an alternative implementation, the weight between the neurons from the perception input layer to the neural computation layer is , and the weight from the neural computation layer to the output control layer is .

[0018] As an alternative implementation, random connection weights are introduced between the neurons of each layer of the shallow brain model to simulate the sparse connection and redundant structure between neurons in the biological nervous system, improving the network robustness.

[0019] As a further limitation, the implementation process of the shallow brain model for controlling the bionic fish swimming is as follows:

[0020] Step 101: Distance sensors are installed at different positions of the robotic fish. Through ultrasonic technology, the relative distance between the robotic fish and surrounding obstacles is measured in real time.

[0021] As an alternative implementation, the distance sensors can be ultrasonic sensors, laser sensors, infrared sensors, etc. When using ultrasonic technology, the distance calculation is expressed as:

[0022]

[0023] where is the straight-line distance between the biomimetic robotic fish and the obstacle, is the round-trip time of the signal, is the propagation speed of the sound wave.

[0024] Step 102: The angle sensor collects the relative angle information between the robotic fish and the obstacle, and updates the angle between the obstacle and the robotic fish in real time according to the motion state of the robotic fish and environmental conditions.

[0025] Step 103: The data is input into the hippocampus module and the cerebellum module. The hippocampus module processes the distance data and generates corresponding perceptual outputs; the cerebellum module processes the angle data, calculates the motion direction of the robotic fish, and adjusts the motion strategy.

[0026] As an alternative implementation, the Tansig activation function is used for data processing in the neural network. Through its non-linear characteristics, it helps the neural network process complex non-linear decision-making tasks.

[0027]

[0028] Tansig is a hyperbolic tangent function that maps the output data to where u is the independent variable of the activation function.

[0029] Step 104: The hippocampus module and the cerebellum module in the neural computing layer calculate independently, and integrate their respective output information to generate the final motion control signal.

[0030] As an alternative implementation, the specific process from the input to the output of the shallow brain model is as follows:

[0031] Step 201: The input to the hippocampus is the Euclidean distance between the robotic fish and the obstacle obtained through the distance sensor:

[0032]

[0033] The difference in the horizontal direction is and the distance in the vertical direction is , the square root of the sum of the squared differences is calculated to obtain the Euclidean distance.

[0034] Step 202: The bionic robotic fish simulates the function of the cerebellum and dynamically adjusts its movement trajectory and direction according to real-time angular perception data, which can be expressed as:

[0035]

[0036] The angle is in radians and represents the direction from the first point to the second point. and respectively represent the differences in positions of the robotic fish and the obstacle on the axis and the axis.

[0037] Step 203: Sense the data transfer between the input layer and the neural computing layer:

[0038]

[0039] For a single neuron in the hidden layer , its input is the distance and angle data received by the sensor. The weight is the weight parameter between the th neuron in the input layer and the th neuron in the hidden layer. The bias is , is the activation function, is the output of the hidden layer.

[0040] Step 204: The data transfer between the neural computing layer and the output control layer is:

[0041]

[0042] Among them, is the number of output layers, is the weight between the th neuron in the output layer and the th neuron in the hidden layer, is the bias term of the th neuron in the output layer. is the final output result.

[0043] The present invention constructs a shallow brain model and controls the swimming of a bionic fish based on the shallow brain model, enabling the bionic fish to avoid obstacles autonomously. The shallow brain model provides an efficient control and computing framework by simplifying and simulating the lower-level neural activities in the biological nervous system. Generally, neurons are divided into three different levels, including a sensory input layer, a neural computing layer, and an output control layer. Neurons at each level are responsible for processing different types of information and interacting with neurons in other levels. In addition, the connections between neurons are randomized, that is, the connection relationships between neurons are not pre-set, but are randomly generated based on a certain probability, breaking the limitation of full connection in traditional neural networks, enabling the network to dynamically adjust connections in different environments, thereby increasing the adaptability and robustness of the network. The sparse connections between neurons reduce the number of connections between neurons, lowering the computational complexity and memory requirements. The neural network adopts a recursive fractal structure, and neurons at each level are connected to the previous level recursively, having similar fractal characteristics in structure, capable of performing fast calculations and processing through similar structures.

[0044] The beneficial effects of the present invention are as follows:

[0045] The present invention provides a method for controlling the swimming of a bionic fish based on a shallow brain model. By simulating the working principle of the biological nervous system, the perception and adaptive obstacle avoidance of the bionic fish are realized. Through the recursive fractal neural network structure, the shallow brain model can efficiently process data from different sensors and quickly generate motion control signals, enabling the robotic fish to flexibly adjust its motion direction to avoid obstacles in a complex environment. Secondly, the shallow brain model has good generalization ability and strong adaptability, and can maintain good motion control performance in a dynamically changing environment. In addition, the randomized connections of neurons enhance the fault tolerance of the network, enabling the system to avoid falling into local optimal solutions, thereby improving the overall robustness. BRIEF DESCRIPTION OF THE DRAWINGS

[0046] The accompanying drawings forming a part of this specification are used to provide a further understanding of the present invention. The schematic embodiments of the present invention and their descriptions are used to explain the present invention and do not constitute an improper limitation of the present invention.

[0047] Figure 1 It is a block diagram of the shallow brain model structure for at least one embodiment of the present invention;

[0048] Figure 2 It is a schematic diagram of information processing of the shallow brain model for at least one embodiment of the present invention;

[0049] Figure 3 It is a schematic diagram of the shallow brain model controlling the bionic fish for at least one embodiment of the present invention; DETAILED DESCRIPTION OF THE EMBODIMENTS

[0050] The present invention will be further described below in conjunction with the accompanying drawings and embodiments.

[0051] It should be noted that the following detailed description is exemplary and is intended to provide further explanation of the present invention. Unless otherwise specified, all technical and scientific terms used herein have the same meaning as commonly understood by those of ordinary skill in the technical field to which the present invention belongs.

[0052] It should be noted that the terms used herein are only for describing specific embodiments and are not intended to limit the exemplary embodiments according to the present invention. As used herein, unless the context clearly indicates otherwise, the singular forms are also intended to include the plural forms. In addition, it should be understood that when the terms "comprising" and / or "including" are used in this specification, they indicate the presence of features, steps, operations, devices, components, and / or combinations thereof.

[0053] A method for controlling the swimming of a bionic robotic fish based on a shallow brain model includes the following steps:

[0054] S1: Construct a shallow brain model, which includes a perception input layer, a neural computing layer, and an output control layer. The neural computing layer mainly includes a hippocampus module and a cerebellum module;

[0055] S2: The perception input layer is used to receive the input data from sensors. The sensors include at least one distance sensor and one angle sensor. The input data includes the relative distance information and angle information between the robotic fish and the surrounding environment;

[0056] S3: The neural computing layer mimics the basic working mechanisms of certain specific regions of the brain, and constructs the hippocampus module and the cerebellum module according to the brain functional partition, especially for performing specific tasks such as perception, decision-making, and motion control;

[0057] S4: Simulate the neuron connection mode of real organisms, randomly connect neurons between each layer, and neurons within the layer do not interfere with each other.

[0058] S5: The output control layer generates a motion instruction for the robotic fish according to the control signal generated by the neural computing layer, and drives the tail fin and pectoral fin of the robotic fish to perform swimming actions, realizing swimming control and obstacle avoidance.

[0059] S6: The shallow brain model is used to simulate the swimming control of the bionic robotic fish to obtain the motion trajectory of the bionic robotic fish.

[0060] Specifically, the shallow brain neural network model described in step S1 includes a perception input layer, a neural computing layer, and an output control layer. The model is a recursive classification structure, and each layer of the network or module has a certain self-similarity in structure.

[0061] Specifically, the implementation process of the sensor and the bionic fish swimming control described in step S2 is as follows:

[0062] Step 101: Distance sensors are installed at different positions of the robotic fish, and the relative distance between the robotic fish and surrounding obstacles is measured in real time through ultrasonic technology.

[0063] As an alternative implementation, the distance sensor can be an ultrasonic sensor, a laser sensor, an infrared sensor, etc. The distance calculation using ultrasonic technology is expressed as:

[0064]

[0065] Where, is the straight-line distance between the bionic fish and the obstacle, is the round-trip time of the signal, is the propagation speed of the sound wave.

[0066] Step 102: The angle sensor collects the relative angle information between the robotic fish and the obstacle, and updates the angle between the obstacle and the robotic fish in real time according to the motion state of the robotic fish and environmental conditions.

[0067] Step 103: The data is input into the hippocampus module and the cerebellum module. The hippocampus module processes the distance data and generates the corresponding perceptual output; the cerebellum module processes the angle data and calculates the motion direction adjustment strategy of the robotic fish.

[0068] As an alternative implementation, the Tansig activation function is used for data processing in the neural network. Through its non-linear characteristics, it helps the neural network process complex non-linear decision-making tasks.

[0069]

[0070] Tansig is a hyperbolic tangent function that maps the output data to where u is the independent variable of the activation function.

[0071] Step 104: The hippocampus module and the cerebellum module in the neural computation layer calculate independently, and integrate their respective output information to generate the final motion control signal.

[0072] Specifically, the basic working mechanism of simulating certain specific regions described in step S3 is to model according to the brain functional partition, and build the hippocampus module and the cerebellum module. Other modules, such as the basal ganglia and thalamus, can be added continuously to make the shallow brain model more perfect.

[0073] Specifically, the neurons between each layer of the shallow brain neural network model are randomly connected with a certain probability. The random connection is sparse, which reduces redundant calculations, avoids overfitting, and improves efficiency.

[0074] Specifically, the specific process from the input to the output of the shallow brain model described in step S5 is as follows:

[0075] Step 201: The input to the hippocampus is the Euclidean distance between the robotic fish and the obstacle obtained through the distance sensor:

[0076]

[0077] The difference in the horizontal direction is and the distance in the vertical direction is The square root of the sum of the squares of the differences is calculated to obtain the Euclidean distance.

[0078] Step 202: By simulating the function of the cerebellum, the bionic robotic fish dynamically adjusts its movement trajectory and direction according to real-time angle perception data, which can be expressed as:

[0079]

[0080] The angle is in radians and represents the direction from the first point to the second point. and represent the differences in positions of the robotic fish and the obstacle on the axis and the axis respectively.

[0081] Step 203: Data transfer between the perception input layer and the neural computing layer:

[0082]

[0083] For a single neuron in the hidden layer its input is the distance and angle data received by the sensor. The weight is the weight parameter between the th neuron in the input layer and the th neuron in the hidden layer. The bias is , is the activation function, is the output of the hidden layer.

[0084] Step 204: The data transfer between the neural computing layer and the output control layer is:

[0085]

[0086] where, is the number of output layers, is the weight between the output layer neuron and the hidden layer neuron ​ is the bias term of the output layer neuron . is the final output result

Claims

1. A swimming control method for a bionic robotic fish based on a shallow brain model, characterized in that: Including the following steps: Construct a shallow brain model to imitate the basic working mechanisms of certain specific regions of the brain, model according to brain functional partitions, especially for performing specific tasks such as perception, decision-making, and motion control; The shallow brain neural network model includes a perception input layer, a neural computing layer, and an output control layer; Construct a shallow brain model, including a perception input layer, a neural computing layer, and an output control layer. The neural computing layer mainly includes a hippocampus module and a cerebellum module; The perception input layer is used to receive input data from sensors. The sensors include at least one distance sensor and one angle sensor. The input data includes the relative distance information and angle information between the robotic fish and the surrounding environment; The neural computing layer imitates the basic working mechanisms of certain specific regions of the brain, partitions according to brain function, especially for performing specific tasks such as perception, decision-making, and motion control, constructs a hippocampus module and a cerebellum module, perceives the input data and generates output signals and motion control signals; Simulate the neuron connection mode of real organisms, randomly connect neurons between each layer, and neurons within the layer do not interfere with each other; The output control layer generates a motion instruction for the robotic fish according to the control signal generated by the neural computing layer, drives the tail fin and pectoral fins of the robotic fish to perform swimming actions, and realizes swimming control and obstacle avoidance; The shallow brain model is used to simulate the swimming control of the bionic robotic fish to obtain the motion trajectory of the bionic robotic fish.

2. The biomimetic robot fish swimming control method based on a shallow brain model according to claim 1, wherein: The hippocampus module and the cerebellum module in the neural computing layer are implemented through a recursive fractal neural network structure, which enhances the adaptability and computing efficiency of the neural network through recursive levels and self-similarity.

3. The biomimetic robotic fish swimming control method based on a shallow brain model according to claim 1, characterized in that: The hippocampus module processes the input data from the distance sensor using the weighted sum activation function of the feedforward neural network, generates a distance perception signal, and outputs the Euclidean distance between the robotic fish and the obstacle.

4. The bionic robot fish swimming control method based on a shallow brain model according to claim 1, characterized in that: The cerebellum module processes the input data from the angle sensor through a feedforward neural network, and generates a motion direction adjustment strategy for the robotic fish according to the calculated relative angle signal.

5. The bionic robot fish swimming control method based on a shallow brain model according to claim 1, characterized in that: The implementation process of the shallow brain model controlling the swimming control of the bionic robotic fish is as follows: Step 101: Distance sensors are installed at different positions of the robotic fish, and the relative distance between the robotic fish and the surrounding obstacles is measured in real time through ultrasonic technology; Step 102: The angle sensor collects the relative angle information between the robotic fish and the obstacle, and updates the angle between the obstacle and the robotic fish in real time according to the motion state of the robotic fish and environmental conditions; Step 103: The data is input into the hippocampus module and the cerebellum module. The hippocampus module processes the distance data and generates corresponding perceptual outputs; The cerebellum module processes the angle data, calculates the motion direction of the robotic fish, and adjusts the motion strategy; Step 104: The hippocampus module and the cerebellum module in the neural computing layer calculate independently, and integrate their respective output information to generate a final motion control signal.

6. The bionic robot fish swimming control method based on a shallow brain model according to claim 1, characterized in that: The shallow brain model introduces random connection weights between neurons in each layer, simulates the sparse connection and redundant structure between neurons in the biological nervous system, improves the network robustness, and enhances the network adaptability.

7. The bionic robot fish swimming control method based on a shallow brain model according to claim 1, characterized in that: The specific process of the described shallow brain model is as follows: Step 201: The input of the hippocampus is the Euclidean distance between the robotic fish and the obstacle obtained through the distance sensor: ; The difference in the horizontal direction is , and the distance in the vertical direction is . After calculating the sum of the squares of the differences and taking the square root, the Euclidean distance is obtained; Step 202: The bionic robotic fish simulates the function of the cerebellum and dynamically adjusts its movement trajectory and direction according to real-time angular perception data, which can be expressed as: ; The angle is in radians and represents the direction from the first point to the second point. and respectively represent the differences in positions of the robotic fish and the obstacle on the axis and axis. Step 203: Data transfer between the perception input layer and the neural computing layer: ; For a single neuron in the hidden layer , its input is the distance and angle data received by the sensor. The weight is the weight parameter between the th neuron in the input layer and the th neuron in the hidden layer. The bias is , is the activation function, is the output of the hidden layer; Step 204: The data transfer between the neural computing layer and the output control layer is: ; Among them, is the output layer number, is the neuron of the output layer and the weight between the neurons of the hidden layer is the bias term of the neuron of the output layer . is the final output result.

Images