A Biomimetic Sideline Reflection-Based Control Method for Underwater Robots to Avoid Currents and Falls

Abstract

This invention belongs to the field of intelligent control technology for underwater special robots, and discloses a current-resistant and fall-prevention control method for underwater robots based on biomimetic lateral line reflection. The method includes: using a piezoelectric hydrophone array with a fish-like lateral line distribution and a flow velocity probe to collect flow field data; incrementally modulating and encoding the data into a pulse sequence; filtering out the robot's own noise pulses using a pre-trained rotational speed noise transfer function; inputting the data into a pulse neural network to identify turbulent disturbances and instability precursor signals; and directly driving the actuator through a hardware bypass bus to complete current-resistant and fall-prevention control. This approach reduces end-to-end control latency, achieves high accuracy in instability warnings in turbid water, and enables coordinated control of current disturbances and fall prevention, avoiding the risk of loss of control and loss of the underwater robot due to sudden changes in the flow field.

Description

Technical Field

[0001] This invention relates to the field of intelligent control technology for underwater special robots, and in particular to a method for controlling underwater robots to resist current and prevent falls based on biomimetic lateral line reflection. Background Technology

[0002] Underwater wall-mounted robots are core equipment in high-risk underwater operation and maintenance scenarios. Current mainstream control solutions are mostly based on macroscopic sensing devices such as vision and sonar combined with PID closed-loop control, which has three core defects: First, vision / sonar perception is easily affected by water turbidity, and completely fails in turbid water, making it impossible to identify instability risks; second, the end-to-end control latency of the main controller path is generally high, and it cannot respond to millisecond-level turbulent disturbances; third, the sensing system is easily affected by the self-noise generated by the operation of the body's thrusters, and the functions of anti-current disturbance and anti-fall are located in different control links and cannot respond in a coordinated manner.

[0003] The above content is only used to help understand the technical solution of the present invention and does not represent an admission that the above content is prior art. Summary of the Invention

[0004] The main objective of this invention is to provide a current-resistant and fall-prevention control method for underwater robots based on biomimetic side-line reflection, aiming to solve the technical problems of high response delay in current disturbance and fall prevention control of existing underwater wall operation robots, easy interference of sensing accuracy by the self-noise of the robot body, and high risk of instability and fall in turbid water environments.

[0005] To achieve the above objectives, the present invention provides a method for controlling the underwater robot against current and preventing falls based on biomimetic lateral line reflection. The method includes the following steps: The pressure gradient data and velocity vector data of the surrounding flow field are collected in real time by a piezoelectric hydrophone array and a flow velocity probe. The pressure signal corresponding to the pressure gradient data whose pressure change exceeds the preset threshold is encoded into a pulse sequence output by an incremental modulation mechanism. The piezoelectric hydrophone array and flow velocity probe are arranged symmetrically along the fuselage axis with a density of 1 to 2 per square centimeter, mimicking the lateral line distribution of fish. Based on the current operating speed parameters of the robot's thrusters, the self-noise pulse characteristics generated by the robot's operation are predicted using a pre-trained speed noise transfer function. The pulse signals corresponding to the self-noise pulse characteristics are then filtered out in real time using the output pulse sequence to obtain a denoised flow field characteristic pulse sequence. The denoised flow field characteristic pulse sequence is input into a spiking neural network. The spatiotemporal integration characteristics of the neurons are used to identify turbulent disturbance signals and instability precursor signals of sudden drop in adsorption force. The turbulent disturbance mode corresponds to the anti-flow disturbance requirement, and the instability precursor signal corresponds to the anti-fall requirement. When the pulse neural network detects a turbulence disturbance signal, it outputs a reverse thrust compensation command to the vector thruster through the hardware bypass control bus to resist the turbulence disturbance. When it detects an adsorption force instability signal, it outputs a pressurization control command to the adsorption device to prevent falling. All control actions are completed before the body undergoes a displacement greater than the preset value.

[0006] In one embodiment, the incremental modulation mechanism is as follows: converting the pressure signal P(t) into a pulse sequence. The trigger condition for pulse output is that the pressure difference between adjacent sampling times satisfies |P(t)|. k )-P(t k-1 )|≥θ th , where θ th The preset pressure change threshold is set to a value ranging from 0.1 kPa to 0.5 kPa, P(t) k ) and P(t k-1 The pressure values ​​at adjacent sampling times are represented by ( ). The piezoelectric hydrophone array achieves far-field disturbance sensing based on the dipole source pressure model of potential flow theory, and the sensed pressure values ​​satisfy ( ). .

[0007] In one embodiment, the step of filtering out the pulse signal corresponding to the self-noise pulse characteristics in real time using the output pulse sequence includes: The denoised flow field characteristic pulse sequence is defined as follows: ; Where Ω represents the current rotational speed of the thruster, For speed-noise fitting coefficients, This represents the noise propagation delay, with a value ranging from 0.2ms to 0.5ms. Minimize prediction error online The fitting coefficients are updated in real time at a frequency of 1 kHz. To adapt to the self-noise characteristics under different propulsion power.

[0008] In one embodiment, the spiking neural network is deployed on field-programmable gate array hardware, pre-trained based on unsupervised pulse time-dependent plasticity rules, and employs a leakage integral ignition neuron model. The neuron membrane potential satisfies the following dynamic formula: ,in, For membrane potential time constant, This is the resting potential; when the membrane potential... Exceeding the issuance threshold V th At that time, the neuron outputs a pulse and resets the membrane potential; the risk patterns used by the spiking neural network to identify include at least left-side eddy current impact, right-side suction abrupt change, frontal turbulent impact, and sudden drop in adsorption contact force.

[0009] In one embodiment, the reverse thrust compensation command and the boost control command satisfy a mapping relationship: ,in, For the pulse frequency of the output layer of the spiking neural network, To control the gain, the value range is 1.2~2.5. is the motor response time constant.

[0010] In one embodiment, the pressure boosting control command is to increase the voltage of the adsorption device to twice the rated voltage and maintain it for no more than 500ms, so as to achieve a doubling of the normal adsorption force while avoiding overheating and burning of the adsorption unit.

[0011] In one embodiment, the method further includes: When the artificial lateral sensing system detects a sudden pressure change and the inertial measurement unit detects a micro-vibration trend, the highest priority reflection control logic is triggered.

[0012] Furthermore, to achieve the above objectives, this invention also proposes a biomimetic lateral line reflection-based underwater robot anti-current and anti-fall control device. This device is applied to the biomimetic lateral line reflection-based underwater robot anti-current and anti-fall control method described above. The device includes: The biomimetic lateral line sensing and encoding module is used to collect pressure gradient data and flow velocity vector data of the surrounding flow field in real time through a piezoelectric hydrophone array and a flow velocity probe. The incremental modulation mechanism is used to encode the pressure signal corresponding to the pressure gradient data whose pressure change exceeds a preset threshold into a pulse sequence output. The piezoelectric hydrophone array and the flow velocity probe are arranged symmetrically along the fuselage axis with a density of 1 to 2 probes per square centimeter, mimicking the lateral line distribution rules of fish. The self-noise adaptive filtering module is used to predict the self-noise pulse characteristics generated by the robot's operation based on the current operating speed parameters of the robot's thrusters and a pre-trained speed noise transfer function. It then uses the output pulse sequence to filter out the pulse signals corresponding to the self-noise pulse characteristics in real time, thereby obtaining a denoised flow field characteristic pulse sequence. The neural risk pattern recognition module is used to input the denoised flow field feature pulse sequence into a spiking neural network, and identify turbulence disturbance signals and instability precursor signals of sudden drop in adsorption force through the spatiotemporal integration characteristics of neurons. The turbulence disturbance pattern corresponds to the anti-flow disturbance requirement, and the instability precursor signal corresponds to the anti-fall requirement. The hardware bypass reflection control module is used to output reverse thrust compensation commands to the vector thruster through the hardware bypass control bus to resist flow disturbance when the pulse neural network detects turbulence disturbance signals. When it detects adsorption instability signals, it outputs pressurization control commands to the adsorption device to prevent falling. All control actions are completed before the machine body undergoes a displacement greater than the preset value.

[0013] Furthermore, to achieve the above objectives, the present invention also proposes an underwater robot anti-current and anti-fall control device based on biomimetic lateral line reflection. The underwater robot anti-current and anti-fall control device based on biomimetic lateral line reflection includes: a memory, a processor, and an underwater robot anti-current and anti-fall control program based on biomimetic lateral line reflection stored in the memory and executable on the processor. The underwater robot anti-current and anti-fall control program based on biomimetic lateral line reflection is configured to implement the steps of the underwater robot anti-current and anti-fall control method based on biomimetic lateral line reflection described above.

[0014] Furthermore, to achieve the above objectives, the present invention also proposes a storage medium storing an underwater robot anti-current and anti-fall control program based on biomimetic lateral line reflection. When the underwater robot anti-current and anti-fall control program based on biomimetic lateral line reflection is executed by a processor, it implements the steps of the underwater robot anti-current and anti-fall control method based on biomimetic lateral line reflection described above.

[0015] This invention employs a piezoelectric hydrophone array with a fish-like lateral line distribution and a flow velocity probe to collect flow field data. This data is incrementally modulated and encoded into a pulse sequence. A pre-trained rotational speed noise transfer function filters out the robot's own noise pulses. The data is then input into a pulse neural network to identify turbulent disturbances and instability precursor signals. A hardware bypass bus directly drives the actuator to complete anti-current and anti-fall control. This approach reduces end-to-end control latency, achieves high accuracy in instability warnings in turbid water, and enables coordinated control of anti-current disturbances and anti-fall control, avoiding the risk of loss of control and the underwater robot due to sudden changes in the flow field. Attached Figure Description

[0016] Figure 1 This is a flowchart illustrating the first embodiment of the underwater robot anti-current and anti-fall control method based on biomimetic lateral line reflection of the present invention. Figure 2 This is a structural block diagram of the first embodiment of the underwater robot anti-current and anti-fall control device based on biomimetic lateral line reflection of the present invention.

[0017] The realization of the objective, functional features and advantages of the present invention will be further explained in conjunction with the embodiments and with reference to the accompanying drawings. Detailed Implementation

[0018] It should be understood that the specific embodiments described herein are for illustrative purposes only and are not intended to limit the scope of the invention.

[0019] This invention provides a method for controlling underwater robots against currents and falls based on biomimetic lateral line reflection, referring to... Figure 1 , Figure 1This is a flowchart illustrating the first embodiment of a biomimetic lateral line reflection-based underwater robot anti-current and anti-fall control method according to the present invention.

[0020] In this embodiment, the underwater robot anti-current and anti-fall control method based on biomimetic lateral line reflection includes the following steps: Step S10: Real-time acquisition of pressure gradient data and velocity vector data of the surrounding flow field using a piezoelectric hydrophone array and flow velocity probe. The pressure signal corresponding to the pressure gradient data whose pressure change exceeds a preset threshold is encoded into a pulse sequence output using an incremental modulation mechanism.

[0021] In this embodiment, the executing entity is an underwater robot anti-current and anti-fall control device based on biomimetic lateral line reflection. This underwater robot anti-current and anti-fall control device based on biomimetic lateral line reflection has functions such as data processing, data communication, and program execution. The underwater robot anti-current and anti-fall control device based on biomimetic lateral line reflection can be a computer terminal device or other network device, or other devices with similar functions. This embodiment does not limit the scope of such devices.

[0022] It should be noted that underwater wall-mounted robots are core equipment in high-risk underwater operation and maintenance scenarios. Current mainstream control solutions are mostly based on macroscopic sensing devices such as vision and sonar combined with PID closed-loop control, which has three core defects: First, vision / sonar perception is easily affected by water turbidity, and completely fails in turbid water, making it impossible to identify instability risks; second, the end-to-end control latency of the main controller path is generally high, and it cannot respond to millisecond-level turbulent disturbances; third, the sensing system is easily affected by the self-noise generated by the operation of the body's thrusters, and the functions of anti-current disturbance and anti-fall are located in different control links and cannot respond in a coordinated manner.

[0023] To address the aforementioned technical challenges, this embodiment employs a piezoelectric hydrophone array with a fish-like lateral line distribution and a flow velocity probe to collect flow field data. This data is incrementally modulated and encoded into a pulse sequence. A pre-trained rotational speed noise transfer function filters out inherent noise pulses. The data is then input into a pulse neural network to identify turbulent disturbances and instability precursor signals. A hardware bypass bus directly drives the actuator to achieve anti-current and anti-fall control. This approach reduces end-to-end control latency, provides high accuracy in instability warnings in turbid water, and enables coordinated control of anti-current disturbances and anti-fall control, avoiding the risk of loss of control and the underwater robot due to sudden changes in the flow field. Specifically, it can be implemented as follows.

[0024] In this specific implementation, the piezoelectric hydrophone array and flow velocity probe in this embodiment are arranged symmetrically along the robot's axial direction, mimicking the lateral line distribution pattern of fish, with a density of 1-2 probes per square centimeter. For example, miniature piezoelectric hydrophones (sensing water pressure changes) and flow velocity probes (sensing flow velocity vectors) are distributed on the robot's outer shell. This simulates the lateral line holes of fish, sensing changes in the pressure gradient of the surrounding flow field.

[0025] In one embodiment, the incremental modulation mechanism is as follows: the pressure signal P(t) is converted into a pulse sequence. The trigger condition for pulse output is that the pressure difference between adjacent sampling times satisfies |P(t)|. k )-P(t k-1 )|≥θ th , where θ th The preset pressure change threshold is set to a value ranging from 0.1 kPa to 0.5 kPa, P(t) k ) and P(t k-1 The pressure values ​​at adjacent sampling times are denoted as . Furthermore, the piezoelectric hydrophone array achieves far-field disturbance sensing based on a dipole source pressure model using potential flow theory, and the sensed pressure values ​​satisfy . .

[0026] Step S20: Based on the current operating speed parameters of the robot's thrusters, predict the self-noise pulse characteristics generated by the robot's operation using a pre-trained speed noise transfer function, and use the output pulse sequence to filter out the pulse signals corresponding to the self-noise pulse characteristics in real time to obtain the denoised flow field characteristic pulse sequence.

[0027] In practical implementation, the vibrations and water flow disturbances generated by the underwater robot's thrusters create strong self-noise interference, accounting for more than 60% of the original sensing signal. Traditional filtering schemes cannot adapt to the dynamic noise characteristics at different speeds, easily leading to the incorrect filtering out of effective flow field signals or the retention of a large amount of noise. To solve this problem, this embodiment uses a pre-trained rotational speed noise transfer function to achieve dynamic noise filtering. Before the robot leaves the factory, it is calibrated in a still water environment to test the amplitude, frequency, and time interval characteristics of self-noise pulses at different thruster speeds, generating a rotational speed-self-noise pulse feature mapping table, which is the entity content of the rotational speed noise transfer function.

[0028] Based on the current operating speed parameters of the robot's thrusters, the process of predicting the self-noise pulse characteristics generated by the robot's operation using a pre-trained speed noise transfer function is as follows: Assuming the robot's rated thruster speed is 3000 rpm, the characteristics of the pre-trained speed noise transfer function corresponding to typical speeds are as follows: When the thruster speed is 800 rpm (low-power cruise state), the predicted self-noise characteristics are: amplitude range 10~15mV, pulse interval 20±2ms, duty cycle 10%; when the thruster speed is 1500 rpm (medium-power anti-flow state), the predicted self-noise characteristics are: amplitude range 25~35mV, pulse interval 10±1ms, duty cycle 22%; when the thruster speed is 2400 rpm (high-power maneuver state), the predicted self-noise characteristics are: amplitude range 40~55mV, pulse interval 5±0.5ms, duty cycle 38%. In actual operation, if the current thruster speed is 1500 rpm, the fuselage tilt is 15°, and the water depth is 20m, the corresponding dynamic correction coefficient α=1.05. The final predicted self-noise characteristics are corrected as follows: amplitude range 26.25~36.75mV, pulse interval 10.5±1.05ms, duty cycle 23.1%. The subsequent pulse matching process will only identify pulses that meet this characteristic range as self-noise pulses. The matching accuracy reaches 98%, the self-noise filtering efficiency is 96.2%, and the effective flow field signal error filtering rate is less than 0.3%.

[0029] Furthermore, the pulse signal corresponding to the self-noise pulse characteristics is filtered out in real time using the output pulse sequence, including: The denoised flow field characteristic pulse sequence is defined as follows: Where Ω is the current rotational speed of the thruster, For speed-noise fitting coefficients, The noise propagation delay is taken as a value ranging from 0.2ms to 0.5ms; the prediction error is minimized online. The fitting coefficients are updated in real time at a frequency of 1 kHz. To adapt to the self-noise characteristics under different propulsion power.

[0030] In practical applications, such as this embodiment, the robot's resistance to disturbances is simulated in high-velocity, high-turbulence environments. Environmental characteristics: The unit operates at low load, with residual leakage at 2 m / s within the pipe, accompanied by randomly detached strong Karman vortex streets. The sensing process involves the robot adhering to the tailrace pipe wall (5 m diameter, concrete lining), and a side-line array (64 piezoelectric nodes) monitoring the surrounding flow field in real time. The event trigger is when the left-side sensor array detects a sharp drop in pressure (…). According to the fluid dynamics model, a strong vortex with a diameter of approximately 0.5 m is predicted to be approaching at a relative velocity of 1.5 m / s. Signal processing involves filtering out propeller noise from the signal before inputting it into the neuromorphic chip. The spiking neural network identifies the "left-side suction enhancement" pattern within 1.2 ms.

[0031] Step S30: Input the denoised flow field characteristic pulse sequence into the spiking neural network, and identify the turbulence disturbance signal and the instability precursor signal of the sudden drop in adsorption force through the spatiotemporal integration characteristics of the neurons.

[0032] In practical implementation, traditional artificial neural networks require frame-by-frame sampling and feature extraction when processing flow field sequence signals, resulting in processing delays generally exceeding 10ms, which cannot meet real-time response requirements. Furthermore, they can only extract single-dimensional features in the time or spatial domains, making it difficult to identify the complex spatiotemporal correlation features of turbulence and instability. This embodiment uses a spiking neural network to achieve risk pattern recognition, which can directly process pulse sequence inputs. Utilizing the spatiotemporal integration characteristics of neurons, it simultaneously extracts signal features in both time and spatial dimensions, significantly reducing recognition delay. In this embodiment, the spiking neural network is deployed on field-programmable gate array hardware, pre-trained based on unsupervised pulse time-dependent plasticity rules, and employs a leakage integral ignition neuron model. The neuron membrane potential satisfies the following dynamic formula: ,in, For membrane potential time constant, This is the resting potential; when the membrane potential... Exceeding the issuance threshold V th At that time, the neuron outputs a pulse and resets the membrane potential; the risk patterns used by the spiking neural network to identify include at least left-side eddy current impact, right-side suction abrupt change, frontal turbulent impact, and sudden drop in adsorption contact force. Among them, the turbulent disturbance pattern corresponds to the anti-flow disturbance requirement, and the instability precursor signal corresponds to the drop prevention requirement. The single-pattern recognition time is less than 1.5ms, and the recognition accuracy rate is over 99%.

[0033] Step S40: When the pulse neural network detects a turbulence disturbance signal, it outputs a reverse thrust compensation command to the vector thruster through the hardware bypass control bus to achieve anti-turbulence. When it detects an adsorption force instability signal, it outputs a pressurization control command to the adsorption device to prevent falling. All control actions are completed before the body undergoes a displacement greater than the preset value.

[0034] It should be noted that in this embodiment, the reverse thrust compensation command and the boost control command satisfy a mapping relationship: ,in, For the pulse frequency of the output layer of the spiking neural network, To control the gain, the value range is 1.2~2.5. The motor response time constant is defined as follows. The boost control command increases the voltage of the adsorption device to twice the rated voltage, maintaining this increase for no more than 500ms to achieve a doubling of the normal adsorption force while preventing overheating and burnout of the adsorption unit. The preset displacement is 1mm, but it can be adaptively adjusted according to actual accuracy requirements; this embodiment does not impose any limitations on this. This embodiment can still achieve 100% adsorption safety warning in turbid water environments with turbidity exceeding 100 NTU and complete visual failure. When applied to the maintenance of tailrace pipes in hydropower stations, it can control the robot's offset under turbulent flow to within 2mm, improving its anti-disturbance capability by 50 times compared to traditional PID control.

[0035] In one embodiment, when the artificial lateral sensing system detects a sudden pressure change and the inertial measurement unit detects a micro-vibration trend, the highest priority reflective control logic is triggered. Traditional control schemes use a main controller software link for response, resulting in end-to-end control delays generally exceeding 50ms. This prevents control actions from being completed before the robot becomes unstable, leading to excessive robot deviation or even a fall. This embodiment uses a hardware bypass control bus to achieve reflective control, eliminating the need for a main controller software link and significantly reducing control latency.

[0036] In this embodiment, a piezoelectric hydrophone array with a fish-like lateral line distribution and a flow velocity probe are used to collect flow field data. This data is incrementally modulated and encoded into a pulse sequence. A pre-trained rotational speed noise transfer function filters out the robot's own noise pulses. The data is then input into a pulse neural network to identify turbulent disturbances and instability precursor signals. A hardware bypass bus directly drives the actuator to complete anti-current and anti-fall control. This approach reduces end-to-end control latency, provides high accuracy in instability warnings in turbid water, and enables coordinated control of anti-current disturbances and anti-fall control, avoiding the risk of loss of control and the underwater robot due to sudden changes in the flow field.

[0037] Furthermore, this embodiment of the invention also proposes a storage medium storing an underwater robot anti-current and anti-fall control program based on biomimetic lateral line reflection. When the underwater robot anti-current and anti-fall control program based on biomimetic lateral line reflection is executed by a processor, it implements the steps of the underwater robot anti-current and anti-fall control method based on biomimetic lateral line reflection described above.

[0038] Reference Figure 2 , Figure 2 This is a structural block diagram of the first embodiment of the underwater robot anti-current and anti-fall control device based on biomimetic lateral line reflection of the present invention.

[0039] like Figure 2 As shown, the underwater robot anti-current and anti-fall control device based on biomimetic lateral line reflection proposed in this embodiment of the invention includes: The biomimetic lateral line sensing and encoding module 10 is used to collect pressure gradient data and flow velocity vector data of the surrounding flow field in real time through a piezoelectric hydrophone array and a flow velocity probe. The pressure signal corresponding to the pressure gradient data whose pressure change exceeds a preset threshold is encoded into a pulse sequence output using an incremental modulation mechanism. The piezoelectric hydrophone array and the flow velocity probe are arranged symmetrically along the fuselage axis with a density of 1 to 2 probes per square centimeter, mimicking the lateral line distribution rules of fish. The self-noise adaptive filtering module 20 is used to predict the self-noise pulse characteristics generated by the robot's operation based on the current operating speed parameters of the robot's thrusters through a pre-trained speed noise transfer function, and to filter out the pulse signals corresponding to the self-noise pulse characteristics in real time using the output pulse sequence to obtain the denoised flow field characteristic pulse sequence. The neural risk pattern recognition module 30 is used to input the denoised flow field feature pulse sequence into the spiking neural network and identify turbulence disturbance signals and instability precursor signals of sudden drop in adsorption force through the spatiotemporal integration characteristics of neurons. The turbulence disturbance pattern corresponds to the anti-flow disturbance requirement, and the instability precursor signal corresponds to the anti-fall requirement. The hardware bypass reflection control module 40 is used to output a reverse thrust compensation command to the vector thruster through the hardware bypass control bus to resist flow disturbance when the pulse neural network detects the turbulence disturbance signal. When the adsorption force instability signal is detected, it outputs a pressurization control command to the adsorption device to prevent falling. All control actions are completed before the body undergoes a displacement greater than the preset value.

[0040] In this embodiment, a piezoelectric hydrophone array with a fish-like lateral line distribution and a flow velocity probe are used to collect flow field data. This data is incrementally modulated and encoded into a pulse sequence. A pre-trained rotational speed noise transfer function filters out the robot's own noise pulses. The data is then input into a pulse neural network to identify turbulent disturbances and instability precursor signals. A hardware bypass bus directly drives the actuator to complete anti-current and anti-fall control. This approach reduces end-to-end control latency, provides high accuracy in instability warnings in turbid water, and enables coordinated control of anti-current disturbances and anti-fall control, avoiding the risk of loss of control and the underwater robot due to sudden changes in the flow field.

[0041] This application also provides an underwater robot anti-current and anti-fall control device based on biomimetic lateral line reflection, including a processor, a communication interface, a memory, and a communication bus. The processor, communication interface, and memory communicate with each other through the communication bus. The memory is used to store the underwater robot anti-current and anti-fall control program based on biomimetic lateral line reflection. When the processor executes the program stored in the memory, it implements the above-mentioned underwater robot anti-current and anti-fall control method based on biomimetic lateral line reflection.

[0042] The communication bus mentioned in the aforementioned biomimetic lateral line reflection-based underwater robot anti-current and anti-fall control device can be a Peripheral Component Interconnect (PCI) bus or an Extended Industry Standard Architecture (EISA) bus, etc. This communication bus can be divided into address bus, data bus, control bus, etc.

[0043] The communication interface is used for communication between the aforementioned biomimetic lateral line reflection-based underwater robot anti-current and anti-fall control device and other devices.

[0044] The memory may include random access memory (RAM) or non-volatile memory (NVM), such as at least one disk storage device. Optionally, the memory may also be at least one storage device located remotely from the aforementioned processor.

[0045] The processors mentioned above can be general-purpose processors, including central processing units (CPUs), network processors (NPs), etc.; they can also be digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, or discrete hardware components.

[0046] In the above embodiments, implementation can be achieved entirely or partially through software, hardware, firmware, or any combination thereof. When implemented using software, it can be implemented entirely or partially in the form of a computer program product. The computer program product includes one or more computer instructions. When the computer program instructions are loaded and executed on a computer, all or part of the processes or functions described in the embodiments of this application are generated. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another. For example, the computer instructions can be transmitted from one website, computer, server, or data center to another website, computer, server, or data center via wired (e.g., coaxial cable, fiber optic, digital subscriber line (DSL)) or wireless (e.g., infrared, wireless, microwave, etc.) means. The computer-readable storage medium can be any available medium that a computer can access or a data storage device such as a server or data center that integrates one or more available media. The available medium can be a magnetic medium (e.g., floppy disk, hard disk, magnetic tape), an optical medium (e.g., DVD), or a semiconductor medium (e.g., solid state disk (SSD)).

[0047] It should be noted that, in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitations, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes said element.

[0048] The various embodiments in this specification are described in a related manner. Similar or identical parts between embodiments can be referred to mutually. Each embodiment focuses on describing the differences from other embodiments. In particular, the system embodiments are basically similar to the method embodiments, so the description is relatively simple; relevant parts can be referred to the descriptions of the method embodiments.

[0049] The above embodiments are only used to illustrate the technical solutions of the present invention, and are not intended to limit it. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

[0050] It should be understood that the above are merely illustrative examples and do not constitute any limitation on the technical solutions of the present invention. In specific applications, those skilled in the art can make settings as needed, and the present invention does not impose any restrictions on this.

[0051] It should be noted that the workflow described above is merely illustrative and does not limit the scope of protection of this invention. In practical applications, those skilled in the art can select some or all of the workflow to achieve the purpose of this embodiment according to actual needs, and no restrictions are imposed here.

[0052] In addition, for technical details not described in detail in this embodiment, please refer to the underwater robot anti-current and anti-fall control method based on biomimetic lateral line reflection provided in any embodiment of the present invention, which will not be repeated here.

[0053] Furthermore, it should be noted that, in this document, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or system that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or system. Unless otherwise specified, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or system that includes that element.

[0054] The sequence numbers of the above embodiments of the present invention are for descriptive purposes only and do not represent the superiority or inferiority of the embodiments.

[0055] Through the above description of the embodiments, those skilled in the art can clearly understand that the methods of the above embodiments can be implemented by means of software plus necessary general-purpose hardware platforms. Of course, they can also be implemented by hardware, but in many cases the former is a better implementation method. Based on this understanding, the technical solution of the present invention, or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product is stored in a storage medium (such as read-only memory (ROM) / RAM, magnetic disk, optical disk) and includes several instructions to cause a terminal device (which may be a mobile phone, computer, server, or network device, etc.) to execute the methods described in the various embodiments of the present invention.

[0056] The above are merely preferred embodiments of the present invention and do not limit the scope of the patent. Any equivalent structural or procedural transformations made based on the description and drawings of the present invention, or direct or indirect applications in other related technical fields, are similarly included within the scope of patent protection of the present invention.

[0057] It is understood that the system provided in the embodiments of the present invention corresponds to the method provided in the embodiments of the present invention, and the explanation, examples and beneficial effects of the relevant content can be referred to the corresponding parts of the above methods.

Claims

1. An anti-flow and anti-falling control method for an underwater robot based on bionic side-line reflection, characterized in that, The underwater robot anti-current and anti-fall control method based on biomimetic lateral line reflection includes: The pressure gradient data and velocity vector data of the surrounding flow field are collected in real time by a piezoelectric hydrophone array and a flow velocity probe. The pressure signal corresponding to the pressure gradient data whose pressure change exceeds the preset threshold is encoded into a pulse sequence output by an incremental modulation mechanism. The piezoelectric hydrophone array and flow velocity probe are arranged symmetrically along the fuselage axis with a density of 1 to 2 per square centimeter, mimicking the lateral line distribution of fish. Based on the current operating speed parameters of the robot's thrusters, the self-noise pulse characteristics generated by the robot's operation are predicted using a pre-trained speed noise transfer function. The pulse signals corresponding to the self-noise pulse characteristics are then filtered out in real time using the output pulse sequence to obtain a denoised flow field characteristic pulse sequence. The denoised flow field characteristic pulse sequence is input into a spiking neural network. The spatiotemporal integration characteristics of the neurons are used to identify turbulent disturbance signals and instability precursor signals of sudden drop in adsorption force. The turbulent disturbance mode corresponds to the anti-flow disturbance requirement, and the instability precursor signal corresponds to the anti-fall requirement. When the pulse neural network detects a turbulence disturbance signal, it outputs a reverse thrust compensation command to the vector thruster through the hardware bypass control bus to resist the turbulence disturbance. When it detects an adsorption force instability signal, it outputs a pressurization control command to the adsorption device to prevent falling. All control actions are completed before the body undergoes a displacement greater than the preset value.

2. The underwater robot anti-current and anti-fall control method based on biomimetic lateral line reflection as described in claim 1, characterized in that, The incremental modulation mechanism is as follows: the pressure signal P(t) is converted into a pulse sequence. The trigger condition for pulse output is that the pressure difference between adjacent sampling times satisfies |P(t)|. k )-P(t k-1 )|≥θ th , where θ th The preset pressure change threshold is set to a value ranging from 0.1 kPa to 0.5 kPa, P(t) k ) and P(t k-1 The pressure values ​​at adjacent sampling times are represented by ( ). The piezoelectric hydrophone array achieves far-field disturbance sensing based on the dipole source pressure model of potential flow theory, and the sensed pressure values ​​satisfy ( ). .

3. The underwater robot anti-current and anti-fall control method based on biomimetic lateral line reflection as described in claim 1, characterized in that, The step of filtering out the pulse signal corresponding to the self-noise pulse characteristics in real time using the output pulse sequence includes: The denoised flow field characteristic pulse sequence is defined as follows: ； Where Ω represents the current rotational speed of the thruster, For speed-noise fitting coefficients, This represents the noise propagation delay, with a value ranging from 0.2ms to 0.5ms. Minimize prediction error online The fitting coefficients are updated in real time at a frequency of 1 kHz. To adapt to the self-noise characteristics under different propulsion power.

4. The underwater robot anti-current and anti-fall control method based on biomimetic lateral line reflection as described in claim 1, characterized in that, The spiking neural network is deployed on field-programmable gate array hardware, pre-trained based on unsupervised pulse time-dependent plasticity rules, and adopts a leakage integral ignition neuron model. The neuron membrane potential satisfies the following dynamic formula: ,in, For membrane potential time constant, This is the resting potential; when the membrane potential... Exceeding the issuance threshold V th At that time, the neuron outputs a pulse and resets the membrane potential; the risk patterns used by the spiking neural network to identify include at least left-side eddy current impact, right-side suction abrupt change, frontal turbulent impact, and sudden drop in adsorption contact force.

5. The underwater robot anti-current and anti-fall control method based on biomimetic lateral line reflection as described in claim 1, characterized in that, The reverse thrust compensation command and the boost control command satisfy a mapping relationship: ,in, For the pulse frequency of the output layer of the spiking neural network, To control the gain, the value range is 1.2~2.

5. is the motor response time constant.

6. The underwater robot anti-current and anti-fall control method based on biomimetic lateral line reflection as described in claim 1, characterized in that, The pressure boosting control command increases the voltage of the adsorption device to twice the rated voltage and maintains it for no more than 500ms, so as to double the normal adsorption force while avoiding overheating and burning of the adsorption unit.

7. The underwater robot anti-current and anti-fall control method based on biomimetic lateral line reflection as described in claim 1, characterized in that, The method further includes: When the artificial lateral sensing system detects a sudden pressure change and the inertial measurement unit detects a micro-vibration trend, the highest priority reflection control logic is triggered.

8. A current-resistant and fall-prevention control device for underwater robots based on biomimetic lateral line reflection, characterized in that, The biomimetic lateral reflection-based underwater robot anti-current and anti-fall control device is applied to the biomimetic lateral reflection-based underwater robot anti-current and anti-fall control method as described in any one of claims 1 to 7, the device comprising: The biomimetic lateral line sensing and encoding module is used to collect pressure gradient data and flow velocity vector data of the surrounding flow field in real time through a piezoelectric hydrophone array and a flow velocity probe. The incremental modulation mechanism is used to encode the pressure signal corresponding to the pressure gradient data whose pressure change exceeds a preset threshold into a pulse sequence output. The piezoelectric hydrophone array and the flow velocity probe are arranged symmetrically along the fuselage axis with a density of 1 to 2 probes per square centimeter, mimicking the lateral line distribution rules of fish. The self-noise adaptive filtering module is used to predict the self-noise pulse characteristics generated by the robot's operation based on the current operating speed parameters of the robot's thrusters and a pre-trained speed noise transfer function. It then uses the output pulse sequence to filter out the pulse signals corresponding to the self-noise pulse characteristics in real time, thereby obtaining a denoised flow field characteristic pulse sequence. The neural risk pattern recognition module is used to input the denoised flow field feature pulse sequence into a spiking neural network, and identify turbulence disturbance signals and instability precursor signals of sudden drop in adsorption force through the spatiotemporal integration characteristics of neurons. The turbulence disturbance pattern corresponds to the anti-flow disturbance requirement, and the instability precursor signal corresponds to the anti-fall requirement. The hardware bypass reflection control module is used to output reverse thrust compensation commands to the vector thruster through the hardware bypass control bus to resist flow disturbance when the pulse neural network detects turbulence disturbance signals. When it detects adsorption instability signals, it outputs pressurization control commands to the adsorption device to prevent falling. All control actions are completed before the machine body undergoes a displacement greater than the preset value.

9. A current-resistant and fall-prevention control device for underwater robots based on biomimetic lateral line reflection, characterized in that, The underwater robot anti-current and anti-fall control device based on biomimetic lateral reflection includes: a memory, a processor, and an underwater robot anti-current and anti-fall control program based on biomimetic lateral reflection stored in the memory and executable on the processor. The underwater robot anti-current and anti-fall control program based on biomimetic lateral reflection is configured to implement the steps of the underwater robot anti-current and anti-fall control method based on biomimetic lateral reflection as described in any one of claims 1 to 7.

10. A storage medium, characterized in that, The storage medium stores an underwater robot anti-current and anti-fall control program based on biomimetic lateral line reflection. When the processor executes the underwater robot anti-current and anti-fall control program based on biomimetic lateral line reflection, it implements the steps of the underwater robot anti-current and anti-fall control method based on biomimetic lateral line reflection as described in any one of claims 1 to 7.

Images