A robot mapping and navigation method based on biomimetic whisker touch

Abstract

The application provides a robot mapping and navigation method based on bionic whisker touch, aiming to solve the problem of failure of traditional vision and laser radar systems in complex and low visibility environments. The specific process includes: initializing and self-checking the bionic whisker sensor to ensure that each module works normally; controlling the whisker movement through the pulse neural network, and relying on the central pattern generator to regulate the movement rhythm of the whisker to realize the touch perception of the surrounding environment; the collected touch data is processed through reconstruction filtering and low-pass filtering to extract the position information of the object, and a complete environment map is generated through the grid map construction algorithm; the A * algorithm is used to plan the path of the generated touch map, calculate the optimal path and perform the navigation task. The system is suitable for robot autonomous navigation in complex environments, and can realize efficient and accurate environment perception and navigation through bionic whisker sensors and neural network technology.

Description

Technical Field

[0001] This invention belongs to the fields of bionics and artificial intelligence. Specifically, it relates to a robot mapping and navigation method based on bionic beard tactile feedback. Background Technology

[0002] With the rapid development of artificial intelligence and robotics, bionic robotics and artificial intelligence have become core research directions in the field of intelligent robotics. Traditional robot navigation systems mainly rely on technologies such as vision and lidar, which acquire visual or lidar data of the environment to achieve environmental perception and path planning. However, these technologies have significant limitations in certain scenarios, especially in complex or confined environments, such as those filled with smoke, dust, insufficient light, or complete darkness. In these situations, the perception capabilities of vision and lidar will decrease significantly or even fail completely. In these cases, robot navigation systems relying solely on vision or lidar may not provide sufficiently accurate environmental information, thus affecting the robot's autonomy and the reliability of task execution.

[0003] Tactile sensing systems are gradually emerging as an effective alternative. Similar to an animal's whiskers, tactile sensing systems can perceive information such as the shape, position, and texture of objects in the environment through physical contact, thus providing environmental perception capabilities completely different from vision and lidar. Tactile sensing systems have shown great potential, especially in extreme environments where vision and lidar cannot function effectively. For example, in darkness, visual perception is almost useless, while tactile sensors can continue to detect and perceive environmental information through physical contact.

[0004] Currently, traditional robot navigation systems typically rely on two main perception technologies: The first is visual perception technology, which acquires visual information about the environment using monocular, binocular, or RGB-D cameras. Computer vision algorithms are then used for image processing, feature extraction, and matching to achieve environmental mapping and navigation. This technology performs excellently in well-lit, unobstructed environments, but its performance significantly degrades in poor lighting conditions or when vision is impaired. The second is lidar perception technology, which uses lidar (LIDAR) to emit laser beams and receive reflected signals to measure the distance and shape of objects in the environment, thus achieving environmental mapping. This technology offers high accuracy and robustness, especially in outdoor or large-scale scenarios. However, its perception capabilities are limited in environments with smoke, dust, or complex geometries.

[0005] Therefore, providing a robot navigation system that can work effectively in complex or constrained environments is a technical problem that urgently needs to be solved by those skilled in the art. Summary of the Invention

[0006] This invention aims to provide a robot mapping and navigation system based on beard-inspired tactile feedback. By simulating the tactile perception mechanism of animal whiskers and combining it with signal processing and other technologies, it achieves efficient autonomous navigation. Its main purpose is to enhance the robot's autonomy and adaptability in complex and low-visibility environments, such as disaster relief, industrial applications, and home environments. This invention aims to solve the following three key problems:

[0007] 1. Addressing the limitations of traditional navigation systems in extreme environments: When conventional perception technologies such as vision or lidar fail in environments such as smoke, darkness, or dust, tactile systems provide a reliable means of perception, ensuring that robots can continuously and effectively perceive their surroundings.

[0008] 2. Improved environmental adaptability and flexibility: Through tactile sensing technology, the system enables adaptive navigation in complex environments, enhancing the robot's real-time response to obstacles and dynamic environments, and improving the flexibility of autonomous navigation.

[0009] 3. Provides a low-power, cost-effective solution: Compared to other high-energy-consuming sensing technologies, such as LiDAR and vision systems, the beard-inspired tactile sensor can significantly reduce energy consumption while maintaining efficient navigation, making it suitable for resource-constrained environments.

[0010] To address the above problems, this invention proposes a robot mapping and navigation system based on beard-inspired tactile feedback. First, a beard-inspired tactile sensor is designed, utilizing a spiking neural network to mimic the sensory mechanism of animal whiskers. It acquires object position information through physical contact, ensuring effective perception in complex environments. Second, tactile signal processing and mapping algorithms are developed. Noise reduction is applied to the sensor data, followed by calculation of contact point coordinates to generate a tactile-based environmental map, improving environmental adaptability. Finally, a low-power autonomous navigation and path planning system is designed, employing the A* algorithm for path planning to ensure the robot selects the optimal path and avoids obstacles in dynamic environments. This system not only overcomes the limitations of traditional navigation but also improves the robot's autonomy and reliability in extreme environments.

[0011] The specific workflow of this invention is as follows:

[0012] Step 1: Start the system. The robot will initialize the bionic beard sensor, and then the system will perform a self-check to confirm that all sensors and modules are in normal working order.

[0013] Step 2: The bionic beard sensor is controlled by a pulse neural network to realize the movement of the beard.

[0014] Step 3: Real-time sensing of surrounding tactile information, control of beard movement based on tactile feedback, and transmission of data to signal processing module.

[0015] Step 4: Preprocess the collected tactile data (including reconstruction filtering and low-pass filtering) to extract the position information of the object.

[0016] Step 5: The processed data is fused to generate a complete environmental map. The map generation process uses a technique similar to SLAM.

[0017] Step 6: After generating the tactile map, the system uses the A* algorithm to calculate the optimal path, and the robot performs navigation tasks based on the path planning results.

[0018] The specific steps are as follows:

[0019] Step 1: System Startup and Sensor Initialization

[0020] When the robot starts up, the initialization of the bionic beard sensor involves zero-point calibration and gain calibration.

[0021] 1. Zero-point calibration: Calibrate the initial state of the sensor so that the sensor output is zero when there is no external interference.

[0022] S calibrated =S measured -ΔS

[0023] Where, S measured It is the signal actually measured by the sensor, and ΔS is the calibration offset.

[0024] 2. Gain calibration: Calibrate the gain coefficient G to ensure that the sensor output is linearly related to the displacement.

[0025] S output =G·x

[0026] Where x is the input physical quantity (displacement), and G is the sensor gain coefficient, which is a constant.

[0027] Step 2: Beard movement control

[0028] The whisker pad, as the primary receiver of sensory information, transmits environmental tactile information to the nervous system, involving the trigeminal ganglion (TG), trigeminal nucleus (TN), and facial nucleus (FN). During this process, the central pattern generator (CPG) autonomously generates rhythmic neural activity and sends signals to the FN, thereby regulating whisker movement.

[0029] Beard movement is controlled by a spiking neural network (SNN), which simulates the firing mechanism of biological neurons. The neuronal membrane potential V(t) is calculated using the following method:

[0030]

[0031] Where C m It is the capacitance on the cell membrane surface, G L E L Leakage conductance, which are parameters inherent to the neuron model itself, I app It is an external input; when the membrane potential V(t) exceeds the threshold V... th When the voltage is -55mV, the neuron emits a pulse signal.

[0032] Beard movement is controlled by the Central Pattern Generator (CPG), which controls the movement of the beard pad muscles. Neurons located in the facial nerve nucleus (FN) send motor commands via the facial nerve. Approximately 80% of the protracting neurons in the FN cause the protraction of a single beard hair, while approximately 20% of the retracting neurons cause the retraction of multiple beard hairs.

[0033] The firing rate of FN neurons determines the angle of beard movement.

[0034] When protractor neurons fire, the whiskers extend forward at an angle of:

[0035] θ pro =R pro *θ max

[0036] When retract neurons fire, the whiskers extend forward at an angle of:

[0037] θ ret =R ret *θ min

[0038] Where R pro R ret The firing rates θ of protractor neurons and retractor neurons are respectively. max and θ min Take π / 3 and -π / 4 respectively.

[0039] Step 3: Environmental Perception

[0040] The biomimetic beard sensor collects environmental information by sensing the force and displacement when an object comes into contact with the beard. The formulas involved include Hooke's Law and the acceleration sensing formula.

[0041] 1. Hooke's Law: The force on a beard is directly proportional to its displacement, which conforms to Hooke's Law.

[0042] F = k·x

[0043] Where F is the force sensed by the beard, k is the stiffness coefficient of the beard, and x is the displacement.

[0044] 2. Acceleration Sensing: When the beard senses the acceleration of an object, it uses Newton's second law:

[0045] F = m·a

[0046] Where F is the force sensed by the beard sensor, m is the mass of the object, and a is the acceleration of the object.

[0047] When the beard senses an object, it activates the contact cells in the TG (transient fasciculus) and connects to the FN (transient fasciculus) neurons through the contact cells in the TN (transient fasciculus). This inhibits the firing of protract neurons and stimulates the firing of retract neurons, causing the beard to retract and reducing the curvature of the beard.

[0048] Step 4: Data Preprocessing and Data Acquisition

[0049] Tactile data preprocessing mainly includes reconstruction filtering and low-pass filtering.

[0050] 1. Reconstruction Filtering:

[0051] This process is equivalent to correcting the Hall sensor's displacements x and y based on the sensor's beard angle input θ, in order to reduce noise and errors, and the final output is x. q y q .

[0052] The filter output f(θ(t)) is obtained through a series of filter coefficients b i The weighted sum of the input signal θ is obtained as follows:

[0053] f(θ(t))=b0·θ(t)+b1·θ(t-1)+b2·θ(t-2)+…+b n ·θ(tn)

[0054] Where θ is the beard angle, and the filter coefficient b i Generate using filter design tools such as the filter design module in MATLAB or SciPy.

[0055] The sensor output signal is obtained by subtracting the filter result from the original signal:

[0056] x q (t)=x(t)-f(θ(t))

[0057] y q (t)=y(t)-f(θ(t))

[0058] 2. Low-pass filter:

[0059] Low-pass filters are typically used to remove high-frequency noise from sensor signals while preserving more stable and important signal characteristics. Here, frequencies below 50Hz are retained, and the formula for calculating the filter's output signal y(t) is:

[0060]

[0061] Where q i and c i The filter coefficients are generated using filter design tools (such as the filter design module in MATLAB or SciPy), where t is the discrete time, ti represents the time step before the i-th time step, and x(t) is the input, i.e., the filtered output x from the previous step. q y q y(t) is the filtered output, denoted as x f y f .

[0062] 3. Data Collection:

[0063] According to x f y f Signal calculation of beard contact force:

[0064]

[0065] macon=(lT)*K

[0066] Where l is the Euclidean distance, T is the contact threshold (usually 0.3), K is the contact gain (usually 5.0), and finally the contact force macon is limited to the range of [0.0, 1.0].

[0067] To reduce the amount of data and high-frequency noise, a simple average is performed on the beard angle signal θ and the contact force signal macon.

[0068] When macon is greater than 0, record the coordinates of the end of the beard at this time as the contact point. Finally, save all contact points as a .csv file.

[0069] Step 5: Environmental Mapping

[0070] At this stage, the system uses a method similar to SLAM (Simultaneous Localization and Mapping) to generate an environmental map using tactile data.

[0071] The processed data is fused using a raster map construction algorithm to generate a complete environmental map. This method first converts sensor data into raster form, then fills in the information in the raster using interpolation and data fusion techniques, ultimately generating a high-resolution environmental map that provides an accurate foundation for subsequent navigation and route planning.

[0072] Step 6: Path Planning

[0073] After generating the haptic map, the system uses the A* algorithm (for details, please refer to http: / / theory.stanford.edu / ~amitp / GameProgramming / AStarComparison.html#the-a-star-algorithm) for path planning and calculates the optimal path.

[0074] The A* algorithm finds the optimal path by evaluating a heuristic function f(n):

[0075] f(n) = g(n) + h(n)

[0076] Where g(n) is the actual cost from the starting point to node n, and h(n) is the heuristically estimated cost from node n to the target point.

[0077] The present invention has the following advantages:

[0078] (1) Adapting to complex environments:

[0079] The beard tactile system can operate effectively in low-visibility and extreme environments, overcoming the limitations of traditional vision and lidar. In environments such as smoke-filled disaster sites and dark underground pipes, the physical contact information provided by tactile sensors is irreplaceable.

[0080] (2) Real-time feedback and response:

[0081] The real-time feedback provided by tactile sensors enables robots to respond quickly to changes in their environment. This real-time responsiveness is crucial for obstacle avoidance and path adjustment, ensuring that robots can complete navigation tasks safely and efficiently.

[0082] (3) Low power consumption and high efficiency:

[0083] Compared to vision and lidar systems, the beard tactile sensor consumes less energy, making it suitable for resource-constrained scenarios. The system's low power consumption allows it to maintain high efficiency during extended operation. Attached Figure Description

[0084] Figure 1 A flowchart of a robot mapping and navigation system based on biomimetic beard tactile feedback

[0085] Figure 2 NRP-based simulated robot

[0086] Figure 3 Contact point scatter plot

[0087] Figure 4 Grid Map

[0088] Figure 5 Navigation using the A* algorithm Detailed Implementation

[0089] To describe the technical solution of the invention in more detail, the invention will be further explained below with reference to the accompanying drawings and examples.

[0090] like Figure 1 As shown, the specific steps of the present invention are as follows:

[0091] 1. Bionic beard structure design

[0092] The biomimetic beard is made of flexible material that allows it to bend slightly upon contact with an object. Sensors embedded in the beard detect this physical deformation and output force information about the beard. Each beard sensor has multiple tactile nodes, thus improving spatial resolution. The choice of flexible material and the structural design of the beard can be adjusted according to the needs of different application scenarios. For example, in scenarios requiring higher tactile precision, the beard can be designed to be more sensitive and detailed.

[0093] 2. Beard movement control

[0094] Beard movement is controlled by a spiking neural network (SNN), which mimics the firing mechanism of biological neurons. Beard movement is generated by a central pattern generator (CPG), a self-oscillating control mechanism independent of the external environment. Motor neurons in the facial nerve nucleus (FN) send movement commands via the facial nerve. The firing rate of FN neurons determines the angle of beard movement; the forward and backward angles are regulated by the firing rates of their respective neurons.

[0095] 3. Tactile Data Processing Methods

[0096] The tactile data processing method is one of the core elements of this invention. Tactile data from the beard sensor first undergoes a preprocessing stage, including reconstruction filtering and low-pass filtering. Then, based on the tactile information, the coordinates of the contact point are calculated to obtain the location of the obstacle. The goal of tactile data processing is to transform information from physical contact into actionable environmental perception data, helping the robot make navigation decisions.

[0097] 4. Haptic Map Generation

[0098] The tactile mapping module is another key component of the system. This module gradually constructs a complete environmental map by fusing tactile data from different locations on the robot. The algorithm employs a SLAM-like approach, but builds the map based on tactile data.

[0099] 5. Haptic navigation algorithm

[0100] Based on the constructed tactile map, the robot uses the A* algorithm for path planning, ensuring that it can move autonomously in complex environments. The design goal of the tactile navigation algorithm is to maximize navigation efficiency and adapt to different environmental changes while ensuring safety.

[0101] The above description of the specific implementation methods of the present invention is only for those skilled in the art to understand the present invention, and is not intended to limit the scope of the present invention.

Claims

1. A robot mapping and navigation method based on biomimetic beard tactile sensation, characterized in that, Includes the following steps: Step 1: Activate the bionic beard sensor; Step 2: The bionic beard sensor is controlled by a pulse neural network to realize the movement of the beard; Step 3: Real-time sensing of surrounding tactile information, control of beard movement based on tactile feedback, and transmission of data to signal processing module; Step 4: Preprocess the collected tactile data, including reconstruction filtering and low-pass filtering, to extract the position information of the object; Step 5: Merge the processed data to generate a complete environmental map; Step 6: After generating the haptic map, the system uses the A* algorithm to calculate the optimal path and executes the navigation task based on the path planning results. Step 2 describes the control of beard movement via a spiking neural network. The specific steps are as follows: The beard pad (WP) acts as the primary receiver of sensory information, transmitting environmental tactile information to the nervous system, involving the trigeminal ganglion (TG), trigeminal nucleus (TN), and facial nucleus (FN). During this process, the central pattern generator (CPG) autonomously generates rhythmic neural activity and sends signals to the FN, thereby regulating beard movement. Beard movement is controlled by a spiking neural network (SNN), which simulates the firing mechanism of biological neurons; input pulses... Calculations based on neuronal membrane potential: ； in It is the capacitance on the cell membrane surface. It is the second leakage conductivity. It is an external input, when the membrane potential Exceeding the threshold When the voltage is -55mV, the neuron emits a pulse signal; Beard movement control is generated by the central pattern generator (CPG), which controls the movement of the beard pad muscles; neurons are located in the facial nerve nucleus (FN) and send movement commands through the facial nerve; 80% of the protracting neurons in the FN cause the protraction of a single beard hair, and 20% of the retracting neurons cause the retraction of multiple beard hairs. The firing rate of FN neurons determines the angle of beard movement; When protractor neurons fire, the whiskers extend forward at an angle of: ； When retract neurons fire, the whiskers extend forward at an angle of: ； in , The firing rates of protracting neurons and retracting neurons are respectively. and Take respectively and- .

2. The robot mapping and navigation system based on bionic beard tactile sensation as described in claim 1, characterized in that, Step 3, which involves real-time sensing of surrounding tactile information and controlling beard movement based on tactile feedback, is detailed below: The bionic beard sensor collects environmental information by sensing the force and displacement when an object comes into contact with the beard; the formulas involved include Hooke's Law and the acceleration sensing formula. (1) Hooke's Law: The force on the beard is proportional to the displacement, which conforms to Hooke's Law: ； in, It is the power of the beard's perception. Let be the stiffness coefficient of the beard. For displacement; (2) Acceleration perception: When the beard senses the acceleration of an object, it uses Newton's second law: ； in, Force sensed by the beard sensor. It is the mass of the object. The acceleration of the object; When the beard senses an object, it activates the contact cells in the TG (transient fasciculus) and connects to the FN (transient fasciculus) neurons through the contact cells in the TN (transient fasciculus). This inhibits the firing of protract neurons and stimulates the firing of retract neurons, causing the beard to retract and reducing the curvature of the beard.

3. The robot mapping and navigation system based on bionic beard tactile sensation as described in claim 1, characterized in that, Step 4 involves reconstruction filtering and low-pass filtering to extract the object's position information, as detailed below: (1) Reconstruction filtering: This process is equivalent to inputting the beard angle from the sensor. Displacement of Hall sensor Corrections are made to reduce noise and errors, and the final output is... ; Filter output Through a series of filter coefficients and input signal The weighted sum is obtained as follows: ； The sensor output signal is obtained by subtracting the filter result from the original signal: ； ； (2) Low-pass filtering: A low-pass filter is used to remove high-frequency noise from sensor signals, retaining frequencies below 50Hz. The filter's output signal... The calculation formula is: ； in and These are the filter coefficients, generated using filter design tools. For discrete time, Indicates the first A time step before The input is the filter output from the previous step. , For the filtered output, it is represented as ; (3) Data collection: according to Signal calculation of beard contact force: ； ； in For Euclidean distance This is the contact threshold, set to 0.

3. This is the contact gain, set to 5.

0. Finally, the contact force... Limited to the range [0.0, 1.0]; when When the value is greater than 0, record the coordinates of the end of the beard at that time as the contact point. Finally, save all contact points as a .csv file.

4. The robot mapping and navigation system based on bionic beard tactile sensation as described in claim 1, characterized in that, Step 5 is as follows: The processed data is fused using a raster map construction algorithm to generate a complete environmental map. This method first converts sensor data into raster form, and then fills the information in the raster with interpolation and data fusion techniques to finally generate an environmental map.

5. A robot mapping and navigation system based on bionic beard tactile sensation as described in claim 1, characterized in that, Step 6 is as follows: After generating the haptic map, the system uses the A* algorithm for path planning to calculate the optimal path; The A* algorithm evaluates a heuristic function. To find the optimal path: ； in, From the starting point to the node The actual cost, For nodes Heuristic cost estimation to the target point.

Images