Intelligent automatic charging bionic fugu robot fish and intelligent charging method

Abstract

The invention provides an intelligent automatically-charged bionic fugu robot fish and an intelligent charging method.The robot fish comprises a fish head, the fish head is connected with a fish body, a driving steering engine, a water suction and drainage bin, a core control panel and a lithium battery are integrated in the fish body, the fish body is connected with a fish tail through the driving steering engine, and the fish tail is connected with the water suction and drainage bin. The top of the fish body and the top of the fish tail are provided with solar charging panels in a placoid scale form, and the solar charging panels, the lithium battery, the energy storage battery and the intelligent charging system jointly form a self-energy-supply system. A cruising function of the robotic fish is completed, the water suction and drainage bin is adopted to change the mass of the robotic fish, the working water area range of the robotic fish is increased, energy consumption is reduced, a DQN algorithm intelligent automatic charging system of a soft attention mechanism is added, deep learning of position information and residual electric quantity information is achieved, and the working efficiency is improved. And finally, the automatic charging system of the bionic robotic fish is more intelligent.

Description

Technical Field

[0001] The present invention belongs to the field of underwater robots, and particularly relates to a bionic pufferfish robot with intelligent automatic charging and an intelligent charging method. Background Art

[0002] At present, the research and development of intelligent bionic robot fish have high application prospects in the intelligent robot industry, and underwater intelligent operation has become the development direction of future intelligent robots. Most underwater robots on the market generally adopt two modes: direct power charging or having a cable without the need for charging. In the first case, the robot may lose power during underwater operation and cannot be recovered, and the recycling cost is too high for long-term operation that requires multiple recycling; in the second case, the flexible operation of the robot underwater cannot be completed, and the operating waters of the underwater robot are greatly restricted. Summary of the Invention

[0003] The purpose of the present invention is to provide a bionic pufferfish robot with intelligent automatic charging and an intelligent charging method, to complete the cruising function of the robot fish, increase the operating waters range of the robot fish, reduce energy consumption, and make the automatic charging system more intelligent.

[0004] The purpose of the present invention is achieved by the following technical solutions:

[0005] A bionic pufferfish robot with intelligent automatic charging, comprising: a fish head, the fish head is connected to the fish body, a driving servo, a suction and drainage chamber, a core control board and a lithium battery are integrated inside the fish body, the fish body is connected to the fish tail through the driving servo, and solar charging panels in the form of dermal denticles are installed on the tops of the fish body and the fish tail. The solar charging panels, the lithium battery, an energy storage battery and an intelligent charging system together constitute a self-power supply system.

[0006] Furthermore, it further includes an intelligent monitoring and data transmission system, and the intelligent monitoring and data transmission system includes intelligent sensors, depth sensors, pressure sensors, intelligent cameras, antennas and communication modules. The intelligent sensors have the function of environmental monitoring regarding temperature, pH value and dissolved oxygen; the intelligent camera and the SD memory card constitute an intelligent shooting system, which can record underwater images; the antennas and the communication modules realize real-time data transmission between the robot fish and the upper computer, and the antennas automatically rise out of the water in shallow water to ensure unobstructed communication.

[0007] Furthermore, two searchlights are installed at the positions of the fish eyes of the fish head, an intelligent camera is installed at the fish mouth part of the fish head, and intelligent sensors are installed at the bottom of the fish body.

[0008] Further, a charging hole, an antenna interface, and a power output hole are provided at the rear of the upper part of the fish body; a square hole is opened in the upper part of the fish body for assembling the internal parts of the fish body; grooves are provided around the square hole, and rubber sealing rings are arranged in the grooves to maintain good watertightness between the square hole and the square cover; the square cover is fastened to the fish body.

[0009] Further, two notches are provided in the vertical direction in the lower half of the fish body. Bolts are used to connect the dorsal fin above and the anal fin below, and the fish tail is connected to the caudal fin.

[0010] Further, when the robotic fish cruises in shallow water, the antenna of the antenna and the communication module rises out of the water surface from the antenna interface, enabling the robotic fish to transmit data to and receive data from the host computer, achieving real-time communication and two-way communication.

[0011] Further, the voltage of the lithium battery combines the position information of the bionic fish and the power information set for the battery to float up for charging, automatically activating the solar charging program. The robotic fish discharges the water in the suction and drainage chamber and floats to the water surface to charge the lithium battery through solar energy; when the voltage reaches the highest voltage of the lithium battery, the robotic fish stops charging, sucks water into the chamber, and dives back underwater to work.

[0012] The present invention may further include:

[0013] An intelligent charging method for the above-mentioned intelligent automatically charging bionic pufferfish robotic fish, the method comprising the following steps:

[0014] The intelligent charging system of the bionic robotic fish is combined with a soft attention mechanism DQN algorithm to realize the intelligent charging of the bionic robotic fish. Combining the position information of the bionic fish and the remaining safe power for the bionic robotic fish to float up for charging set by the battery, the solar charging program is automatically activated. The robotic fish discharges the water in the chamber and floats to the water surface to charge the lithium battery through solar energy; when the voltage reaches the highest voltage of the lithium battery, the robotic fish stops charging and returns to work underwater again; the position information of the bionic robotic fish and the current remaining power information for floating up for charging are continuously learned by the soft attention mechanism DQN algorithm to ultimately achieve the goal of intelligent charging.

[0015] Further, the soft attention mechanism DQN algorithm includes the following steps:

[0016] First, a deep Q network is created, including an input layer, a hidden layer, and an output layer. The output layer predicts the Q values of all possible actions when the bionic robotic fish is working.

[0017] An attention module is added to the network structure. This module is a separate neural network layer used to calculate the attention weights of the input state. The attention mechanism is used to calculate the attention weights of the input state, and these weights represent the degree of attention of the network to different parts of the input state;

[0018] According to the calculated attention weights, the position information h state of the input bionic fish is weighted to highlight the information that is more important for the current decision;

[0019] The forward propagation and backward propagation algorithms are used to update the Q-value prediction of the network for the weighted input state; the experience replay and target network techniques in reinforcement learning are used to train the DQN network, while considering the influence of the attention weights; at each time step, the trained DQN network and the attention mechanism are used to select the optimal action for the bionic fish to float up and charge, execute the action in the environment, collect rewards and new state information, and continuously iterate and update the network parameters to achieve the stability, efficiency and safety of the bionic fish intelligent charging system.

[0020] Furthermore, when the bionic fish is cruising, for the given position S of the fish and the set battery level Q1, deep learning and reinforcement learning are used to set the learning data (s, a). The position information of the bionic fish is provided by a depth sensor. When the bionic fish reaches state s during operation and the current battery level displayed at that time satisfies the setting of formula (1), the bionic fish automatically makes the action a of floating up to charge, where:

[0021] Q1 = Q - hp(1)

[0022] In the formula: Q1 is the remaining safe battery level for the bionic fish to float up and charge, Q is the current battery level, h is the depth during the diving operation before the bionic fish floats up to charge, and p is the power consumed by the average floating height of the bionic fish.

[0023] The beneficial effects of the present invention are as follows:

[0024] The top of the fish body of the present invention is provided with a solar charging panel in the form of placoid scales, which is combined with a deep learning intelligent control charging system to realize the automatic charging of the bionic fish. By installing a solar charging panel in the form of placoid scales on the bionic pufferfish body, replacing the configuration chamber with a suction and drainage chamber and setting an intelligent charging system inside, the recovery cost of the bionic pufferfish is greatly reduced. With the design of the suction and drainage chamber, when charging is required, the water in the chamber is discharged to reduce the weight of the bionic fish itself and reduce the power consumption when the bionic pufferfish floats up to charge. When diving, water is sucked into the drainage chamber to increase the weight of the bionic pufferfish itself and reduce the energy consumption during the diving operation. Using clean energy meets the concept of green and low-carbon, and realizes the cruising function of the bionic pufferfish.

[0025] The DQN algorithm intelligent automatic charging system of the present invention incorporates a soft attention mechanism to achieve deep learning of position information and remaining battery information, ultimately ensuring that the automatic charging system of the bionic robotic fish is more intelligent, safer, more stable, and more efficient.

[0026] The present invention replaces the traditional counterweight chamber with a suction and drainage chamber, enabling the underwater robot to arbitrarily change its own weight, thereby achieving the function of diving at multiple positions, greatly increasing the diving operation water area, and unloading all the weight of the water chamber during floating charging to reduce energy consumption.

[0027] The fish body device of the present invention improves the traditional solar panel into a solar charging panel in the form of grooves and blunt scales, achieving drag reduction and improving propulsion efficiency, making the bionic fish move more smoothly underwater, reducing energy consumption, and improving its natural motion characteristics of simulating organisms.

[0028] The present invention adopts an intelligent charging system. When the battery is low, it automatically drains the water in the water chamber, floats to the water surface to complete solar charging, then dives to the designated water area for operation, and can continuously operate without retrieving the underwater robot, greatly reducing the retrieval cost, reducing the biological disturbance to the working water area, and significantly improving the operation endurance of the underwater robot. BRIEF DESCRIPTION OF THE DRAWINGS

[0029] FIG Figure 1 is a schematic structural diagram of the present invention.

[0030] FIG Figure 2 is a schematic structural diagram of the internal cabin of the fish body of the present invention.

[0031] FIG Figure 3 is a flowchart of adding the DQN algorithm with a soft attention mechanism in the present invention.

[0032] In the drawings: 1. Bionic fish head, 2. Fish body, 3. Fish tail, 4. Intelligent camera, 5. Fish tail fin, 6. Pressure sensor, 7. Intelligent sensor, 8. Driving servo suction, 9. Drainage chamber, 10. Solar charging panel in the form of dermal denticles, 11. Core control board, 12. Lithium battery, 13. Antenna interface, 14. Depth sensor. DETAILED DESCRIPTION OF THE INVENTION

[0033] The present invention will be further described below with reference to the drawings.

[0034] An intelligent automatic charging bionic pufferfish robotic fish, as shown in FIG Figure 1-2As shown in the figure, it includes: a fish head 1, the fish head 1 is connected to the fish body 2, the fish body 2 integrates a driving servo 8, a water suction and drainage chamber 9, a core control board 11 and a lithium battery 12 inside. The fish body 2 is connected to the fish tail 3 through the driving servo 8. On the top of the fish body 2 and the fish tail 3, there is a solar charging panel 10 in the form of dermal denticles. The solar charging panel 10, the lithium battery 12, the energy storage battery and the intelligent charging system together constitute a self-powered system. The driving system composed of the driving servo and the fish tail ensures the efficient movement of the robotic fish; the single-chip microcomputer carried by the core control board is responsible for data processing and the coordinated operation of each component of the robotic fish. The intelligent charging system can sense the remaining battery power and automatically start the charging program to ensure that the robotic fish can achieve self-supplementation of energy both underwater and on the water surface.

[0035] In this embodiment, the seamless combination of the fish head 1 and the fish body 2, the fish head and the fish body are connected by bolts; the fish tail 3 is connected to the fish body through a precise driving system, ensuring the stability and flexibility of the robotic fish. There are two notches in the lower half of the fish body 2 in the up and down direction. The dorsal fin is connected above by bolts, and the anal fin is connected below. The fish tail 3 is connected to the caudal fin 5 of the fish tail. The connection of the dorsal fin and the anal fin through the notches in the lower half of the fish body and the bolts, and the combination of the caudal fin 5 of the fish tail and the fish tail 3 further improve the underwater mobility and stability of the robotic fish.

[0036] The retractable handle on the outside of the fish body 2 simplifies the process of launching and recovering the robotic fish, ensuring the convenience of operation.

[0037] Specifically, the present invention adopts a multi-functional hole and sealing design. The small round holes behind the fish body 1 respectively undertake the functions of charging, antenna connection and power output; the square hole on the top of the fish body 2 is convenient for the assembly of internal components, and the rubber sealing ring ensures the tightness of underwater operation. The square cover fixed by screws ensures the durability and watertight safety of the robotic fish. The groove inside the edge of the square hole is provided with a rubber sealing ring, which fits closely with the square cover to form an effective waterproof barrier. The square cover is fixed on the fish body by screws to ensure the stability and watertightness of the overall structure.

[0038] There are two searchlights installed at the position of the fish eyes of the fish head 1, and an intelligent camera 4 is installed at the fish mouth part of the fish head 1. An intelligent sensor 7 is installed at the bottom of the fish body.

[0039] In this embodiment, there are a charging hole, an antenna interface 12 and a power output hole at the rear of the upper part of the fish body 1; a square hole is opened in the upper part of the fish body 2 for assembling the internal parts of the fish body 2; there is a groove around the square hole, and a rubber sealing ring is arranged in the groove to keep the square hole and the square cover in good watertightness; the square cover is fastened on the fish body 2.

[0040] This embodiment further includes an intelligent monitoring and data transmission system, which includes an intelligent sensor 7, a depth sensor 14, a pressure sensor 6, an intelligent camera 4, an antenna and a communication module. The intelligent sensor 7 has the function of environmental monitoring regarding temperature, pH value, and dissolved oxygen. The intelligent camera 4 and the SD memory card form an intelligent shooting system, which can record underwater images. The antenna and the communication module realize real-time data transmission between the robotic fish and the host computer. The antenna automatically rises out of the water at shallow depths to ensure unobstructed communication. The intelligent sensor 7 is installed at the bottom of the fish body 2. The acquisition depth sensor 14 determines the position and depth of the bionic fish.

[0041] In this embodiment, when the robotic fish cruises at a shallow depth, the antenna of the antenna and communication module rises out of the water from the antenna interface 13, realizing the mutual transmission of data between the robotic fish and the host computer, achieving real-time communication and mutual communication.

[0042] The voltage of the lithium battery 12 combines the position information of the bionic fish and the power information set for floating up and charging of the battery, automatically activating the solar charging program. The robotic fish drains the water in the suction and drainage chamber 9 and floats up to the water surface to charge the lithium battery 12 through solar energy. When the voltage is charged to the maximum voltage of the lithium battery 12, the robotic fish stops charging, sucks water into the chamber, and dives back underwater to work.

[0043] In this embodiment, the suction and drainage chamber can change its own weight to achieve multi-directional diving positions. When floating up for solar charging, the power consumption can be greatly reduced. The solar panel adopts an innovative placoid scale form, which can achieve drag reduction and improve the propulsion efficiency. This structural design can make the bionic fish move more smoothly underwater, reduce energy consumption, and improve its natural motion characteristics of simulating organisms. Equipped with an intelligent charging system, it automatically floats up to charge when the power is low and automatically dives to work when fully charged, greatly reducing the recovery cost. Equipped with a solar charging panel, it follows the concept of green environmental protection.

[0044] According to the design characteristics of the bionic fish of the present invention, it is ensured that the bionic robotic fish will not cause other functions to malfunction due to too low power before floating up for charging. On this basis, combined with the total amount of energy stored in the designed battery, the safety power Q1 is set to 5% of the total battery energy.

[0045] On this basis, this embodiment further includes:

[0046] An intelligent charging method for the above intelligent automatically charging bionic pufferfish robotic fish, the method includes the following steps:

[0047] The intelligent charging system of the bionic robotic fish incorporates a DQN algorithm with a soft attention mechanism to achieve intelligent charging of the bionic robotic fish. By combining the position information of the bionic fish and the battery settings, the remaining safe power for the bionic robotic fish to float up for charging is set, and the solar charging program is automatically activated. The robotic fish drains the water in the water tank and floats to the water surface to charge the lithium battery through solar energy. When the voltage reaches the maximum voltage of the lithium battery, the robotic fish stops charging and returns to work underwater. The position information of the bionic robotic fish and the current remaining power information for floating up and charging are continuously learned using the DQN algorithm with a soft attention mechanism, ultimately achieving the goal of intelligent charging.

[0048] Further, the DQN algorithm with a soft attention mechanism includes the following steps, which will be described in conjunction with the attached Figure 3 for illustration:

[0049] First, create a deep Q-network, including an input layer, a hidden layer, and an output layer. The output layer predicts the Q-values of all possible actions when the bionic robotic fish is working.

[0050] Add an attention module to the network structure. This module is a separate neural network layer used to calculate the attention weights of the input state. The attention mechanism is used to calculate the attention weights of the input state, and these weights represent the degree of attention of the network to different parts of the input state.

[0051] According to the calculated attention weights, weight the input position information h state of the bionic robotic fish to highlight the information that is more important for the current decision.

[0052] Use the forward propagation and backpropagation algorithms to update the Q-value prediction of the network for the weighted input state. Use the experience replay and target network techniques in reinforcement learning to train the DQN network, taking into account the influence of the attention weights at the same time. At each time step, use the trained DQN network and the attention mechanism to select the optimal action for the bionic robotic fish to float up and charge, execute the action in the environment, collect rewards and new state information, and continuously iterate and update the network parameters to achieve the stability, efficiency, and safety of the intelligent charging system of the bionic robotic fish.

[0053] Further, when the bionic robotic fish is cruising, for the given position S of the robotic fish and the set power Q1, deep learning and reinforcement learning are used to set the learning data (s, a). The position information of the bionic robotic fish is provided by a depth sensor. When the bionic robotic fish reaches the s state during work and the current power displayed at that time satisfies the setting of formula (1), the bionic robotic fish automatically makes the action a of floating up for charging, where:

[0054] Q1 = Q - hp(1)

[0055] Where: Q1 is the remaining safe power for the bionic robotic fish to float up for charging, Q is the current battery power, h is the depth during the diving operation of the bionic robotic fish before floating up for charging, and p is the power consumed by the bionic robotic fish for the average floating height.

[0056] In this embodiment, based on the conditional formula and combined with the depth sensor and lithium battery voltage signal sensor possessed by the present invention itself, an improved DQN algorithm is proposed. For the depth h information measured by the depth sensor, a soft attention mechanism is added, and the network assigns a weight to each depth information, and these weights are calculated by an attention network. Then, these weights are used to weighted average the input features to form a context vector, which is subsequently used for the calculation of the Q value (here Q refers to the Query vector). This method allows the network to focus on different input regions during each decision-making. On the basis of introducing the soft attention and combining the above formula, the position information h is associated with the current power Q of the battery. The addition of this mechanism can enable the bionic robotic fish of the present invention to dive to any position for work, so that the DQN network can dynamically focus on different parts of the input state. By assigning different attention weights to the position information h, the network can extract and utilize information more effectively, especially when dealing with long-distance dependence relationships. Combining the formula to link the position information h with the current power Q information realizes a more accurate and stable floating-up charging instruction.

[0057] In this embodiment, Experience replay in the DQN algorithm can set a memory bank to learn from previous experiences, current experiences, and others' experiences, and Fixed Q-targets in the DQN algorithm can disrupt the correlation mechanism for continuous update learning. The intelligent charging system of the bionic robotic fish with the specific soft attention mechanism DQN algorithm of the present invention can realize the intelligent charging of the bionic robotic fish. Combining the position information of the bionic fish and the battery, the remaining safe power for the bionic robotic fish to float up for charging is set, and the solar charging program is automatically activated. The robotic fish discharges the water in the water tank and floats to the water surface to charge the lithium battery through solar energy; when the voltage is charged to the highest voltage of the lithium battery, the robotic fish stops charging and returns to work underwater. The position information of the bionic robotic fish and the current remaining power information for floating up for charging are continuously learned by the improved DQN algorithm to finally achieve the goal of intelligent charging.

[0058] The above are only the preferred embodiments of the present invention and are not intended to limit the present invention. For those skilled in the art, the present invention can have various changes and modifications. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principle of the present invention shall be included within the protection scope of the present invention.

Claims

1. A bionic pufferfish robot fish with intelligent automatic charging, characterized in that, Comprising: A fish head (1), the fish head (1) is connected to the fish body (2), inside the fish body (2), a driving servo (8), a suction and drainage chamber (9), a core control board (11) and a lithium battery (12) are integrated. The fish body (2) is connected to the fish tail (3) through the driving servo (8). On the tops of the fish body (2) and the fish tail (3), solar charging panels (10) in the form of placoid scales are installed. The solar charging panels (10), the lithium battery (12), an energy storage battery and an intelligent charging system together constitute a self-powered system.

2. The bionic pufferfish robot fish with intelligent automatic charging according to claim 1, characterized in that, It further includes an intelligent monitoring and data transmission system. The intelligent monitoring and data transmission system includes an intelligent sensor (7), a depth sensor (14), a pressure sensor (6), an intelligent camera (4), an antenna and a communication module. The intelligent sensor (7) has the environmental monitoring function regarding temperature, pH value and dissolved oxygen; the intelligent camera (4) and an SD memory card constitute an intelligent shooting system, which can record underwater images; the antenna and the communication module realize the real-time data transmission between the robotic fish and the upper computer. The antenna automatically rises out of the water at shallow depths to ensure unobstructed communication.

3. The intelligent and automatically rechargeable bionic pufferfish robot fish according to claim 1 or 2, characterized in that, Based on the voltage of the lithium battery (12), combined with the position information of the bionic fish and the power information set for the battery to float up for charging, the solar charging program is automatically activated. The robotic fish discharges the water in the suction and drainage chamber (9) and floats up to the water surface to charge the lithium battery (12) through solar energy; when the voltage is charged to the maximum voltage of the lithium battery (12), the robotic fish stops charging, sucks water into the water chamber and dives back underwater to work.

4. The bionic pufferfish robot fish for intelligent automatic charging and solar power generation without recycling according to claim 3, wherein At the positions of the fish eyes of the fish head (1), two searchlights are installed. At the fish mouth part of the fish head (1), an intelligent camera (4) is installed. At the bottom of the fish body, an intelligent sensor (7) is installed.

5. The bionic pufferfish robot fish with intelligent automatic charging according to claim 4, characterized in that, At the rear of the upper part of the fish body (1), there are a charging hole, an antenna interface (12) and a power output hole; on the upper part of the fish body (2), a square hole is opened for assembling the internal parts of the fish body (2); rubber sealing rings are arranged in the grooves around the square hole to keep good water tightness between the square hole and the square cover; the square cover is fastened on the fish body (2).

6. The biomimetic pufferfish robot fish with intelligent automatic charging according to claim 5, characterized in that, On the lower half of the fish body (2), two notches are arranged in the up-down direction. The dorsal fin is connected above by bolts, and the anal fin is connected below. The fish tail (3) is connected to the caudal fin (5).

7. The bionic tetraodontiform robotic fish with intelligent automatic charging according to claim 6, characterized in that When the robotic fish cruises at shallow depths, the antenna of the antenna and communication module rises out of the water from the antenna interface (13) to realize the mutual data transmission between the robotic fish and the upper computer, and realize real-time communication and mutual communication.

8. An intelligent charging method for the intelligent automatically charging bionic pufferfish robot fish according to any one of claims 1-7, characterized in that, The method includes the following steps: The intelligent charging system of the bionic robotic fish incorporates a DQN algorithm with a soft attention mechanism to achieve intelligent charging of the bionic robotic fish. By combining the position information of the bionic fish and the battery settings, the remaining safe charge for the bionic robotic fish to float up for charging is determined, automatically activating the solar charging program. The robotic fish drains the water in the water tank and floats to the water surface to charge the lithium battery through solar energy. When the voltage reaches the maximum voltage of the lithium battery, the robotic fish stops charging and returns to work underwater. The position information of the bionic robotic fish and the current remaining charge information for floating up for charging are continuously learned using the DQN algorithm with a soft attention mechanism to ultimately achieve the goal of intelligent charging.

9. The intelligent automatic charging method of the bionic pufferfish family robotic fish for intelligent automatic charging without recycling and solar power generation according to claim 8, characterized in that, The DQN algorithm with a soft attention mechanism includes the following steps: First, create a deep Q-network, including an input layer, hidden layers, and an output layer. The output layer predicts the Q-values of all possible actions when the bionic robotic fish is working. Add an attention module to the network structure. This module is a separate neural network layer used to calculate the attention weights of the input state. The attention mechanism is used to calculate the attention weights of the input state, and these weights represent the degree of attention of the network to different parts of the input state. According to the calculated attention weights, weight the input position information h state of the bionic robotic fish to highlight the information that is more important for the current decision. Use the forward propagation and backward propagation algorithms to update the Q-value prediction of the network for the weighted input state. Use the experience replay and target network techniques in reinforcement learning to train the DQN network, while considering the influence of the attention weights. At each time step, use the trained DQN network and the attention mechanism to select the optimal action for the bionic robotic fish to float up for charging, execute the action in the environment, collect rewards and new state information, and continuously iterate and update the network parameters to achieve the stability, efficiency, and safety of the intelligent charging system of the bionic robotic fish.

10. The intelligent automatic charging method of the bionic pufferfish robot fish with intelligent automatic charging according to claim 9, characterized in that, When the bionic robotic fish is cruising, for a given position S of the robotic fish and a set charge Q1, deep learning and reinforcement learning are used to set the learning data (s, a). The position information of the bionic robotic fish is provided by a depth sensor. When the bionic robotic fish is working and reaches state s and the current charge displayed at that time satisfies the setting of formula (1), the bionic robotic fish automatically makes the action a of floating up for charging, where: Q1 = Q - hp (1) where: Q1 is the remaining safe charge for the bionic robotic fish to float up for charging, Q is the current charge of the battery, h is the depth during the diving operation before the bionic robotic fish floats up for charging, and p is the power consumed per unit height of the bionic robotic fish floating up.

Images