Amphibious snakelike robot

Abstract

The invention relates to the technical field of snakelike robots, and particularly discloses an amphibious snakelike robot which comprises a mechanical structure, a power system, a sensor system and a control and communication system. The mechanical structure comprises a main trunk unit which is a rigid section made of a high-strength lightweight alloy; the flexible connecting units are bendable joints made of tough composite materials and are used for connecting the adjacent main trunk units, and the main trunk units and the flexible connecting units are connected in series to form a snake-like bionic structure and provide basic support and power bearing; the motion executing mechanism comprises a wheel type driving assembly and a propelling assembly; according to the application, by adding the soil, underground water and hydrocarbon substance in-situ sensor system, the'rescue-monitoring 'integrated capability is formed, the environment detection dimension is expanded, 'rescue + risk assessment' dual-mode collaboration is realized, and meanwhile, the detection reliability and data representativeness in a complex environment can also be improved.

Description

TECHNICAL FIELD

[0001] The application belongs to the technical field of snake-shaped robots, and particularly relates to an amphibious snake-shaped robot. BACKGROUND

[0002] The amphibious snake-shaped robot is a kind of bionic robot, which combines the movement characteristics of snakes and amphibious ability, can flexibly switch the action mode on land and in water, and is suitable for complex environments. The core design concept of the amphibious snake-shaped robot is derived from the “limbless movement” mechanism of snake species, and the winding, side shifting, rolling and other actions are realized through a multi-joint modular structure, and special driving devices (such as passive wheels, propellers or air bags) are combined to meet the movement needs in different environments. In recent years, the demand for amphibious robots in disaster rescue, industrial detection, underwater exploration and other fields is increasing.

[0003] In recent years, with the upgrading of the demand in the fields of disaster emergency, industrial detection, environmental governance and the like, the demand for “mobile rescue + environmental monitoring” integrated equipment is increasingly urgent. The technology of traditional rescue robots and environmental detection equipment is fragmented, and it is difficult to meet the core requirements of “fast response, multi-dimensional evaluation and cross-medium operation” in complex scenarios. The specific technical bottlenecks are as follows: ① The environmental monitoring capability of the existing rescue robots is limited. The technical focus of the existing amphibious rescue robots (such as amphibious bionic robots and wheeled rescue robots) is mainly on terrain adaptation and target search and rescue, and the environmental monitoring function has a significant shortcoming. ② Single detection dimension: the mainstream products are only equipped with basic sensors such as vision and radar, focusing on terrain modeling and trapped personnel identification, and lacking the detection capability of soil pollution, groundwater quality, flammable and explosive hydrocarbon chemicals and the like. For example, in the earthquake debris rescue, the robot can locate the trapped personnel but cannot detect the heavy metal leakage in the surrounding soil, which leads to the risk of secondary injury to the rescue personnel; in the flood disaster, the underwater search and rescue robot can only identify the fallen target, but cannot monitor whether the water body is toxic due to chemical leakage, which restricts the comprehensiveness of the rescue decision; ③ Poor cross-medium monitoring adaptability: some amphibious robots try to integrate simple environmental sensors (such as gas detectors), but are not optimized for the “land-underwater” cross-medium characteristics. In the land mode, the sensor is easy to be disturbed by the bumps and dust, leading to data distortion; in the underwater mode, due to the design defects of the sealing, the sensor cannot realize the in-situ real-time collection of water quality parameters, and the detection response time is generally ≥5 seconds, which is far behind the timeliness requirement of the rescue action. SUMMARY

[0004] The purpose of the present application is to provide an amphibious snake-shaped robot to solve the problems raised in the background.

[0005] To achieve the above-mentioned purpose, the present application provides the following technical scheme:

[0006] An amphibious snake-shaped robot comprises:

[0007] mechanical structure, power system, sensor system, control and communication system;

[0008] The mechanical structure comprises:

[0009] a main trunk unit made of a rigid segment of high-strength lightweight alloy;

[0010] a flexible connection unit made of a bendable joint of ductile composite material for connecting adjacent main trunk units, the main trunk unit and the flexible connection unit being connected in series to form a snake-like bionic structure, providing basic support and power bearing;

[0011] a motion execution mechanism, the motion execution mechanism comprising a wheeled driving assembly and a propulsion assembly, the wheeled driving assembly being mounted at the bottom of the main trunk unit, the wheeled driving assembly being rigidly connected with the main trunk unit to realize land movement function, the propulsion assembly being arranged at the terminal trunk unit, the propulsion assembly being connected with the terminal main trunk unit through a quick connector to realize underwater propulsion mode switching.

[0012] Preferably, the power system comprises:

[0013] an energy module and a power transmission system;

[0014] The energy module comprises: a lithium battery pack and an intelligent battery management system, the lithium battery pack being distributed along the axis of the main trunk unit, the intelligent battery management system being electrically connected with each battery pack;

[0015] The power transmission system comprises: a wheeled driving motor and an underwater propulsion motor, the wheeled driving motor being connected with the driving wheel through a reduction gear set, the underwater propulsion motor being connected with the propeller and the bionic fin assembly of the propulsion assembly through a universal joint.

[0016] Preferably, the sensor system comprises:

[0017] an environment perception module, a pose detection module, a special sensor module, a soil parameter in-situ detection module, a groundwater quality in-situ monitoring module and a hydrocarbon in-situ detection module;

[0018] The environment perception module comprises: a high-definition camera, an infrared thermal imager and a laser radar, the high-definition camera and the infrared thermal imager forming a binocular vision unit, the laser radar being mounted at the front end of the robot;

[0019] The pose detection module comprises: a nine-axis IMU and a depth sensor, the nine-axis IMU being internally provided with a gyroscope, an accelerometer and a magnetometer, the depth sensor being arranged at the bottom of the robot;

[0020] The special sensor includes a multi-gas detector and a bioelectric field sensor.

[0021] The high-definition camera and the infrared thermal imager are connected with the first main trunk unit through a holder mechanism to realize three-dimensional reconstruction of the environment, the laser radar is directly connected with the main control system to establish a SLAM environment map, and the nine-axis IMU is fixed at the centroid position of the robot to transmit pose data through an I2C interface.

[0022] Preferably, the soil parameter in-situ detection module includes:

[0023] A soil humidity / conductivity sensor, a soil pH sensor, and a soil heavy metal ion sensor; the soil parameter in-situ detection module is arranged at the bottom of the middle main trunk unit of the robot, a spring-driven telescopic probe is designed, and the protection level reaches IP68;

[0024] The groundwater quality in-situ monitoring module includes:

[0025] A dissolved oxygen sensor, a turbidity sensor, a total organic carbon sensor, and a heavy metal rapid detector, and the groundwater quality in-situ monitoring module is integrated in a waterproof detection cabin outside the terminal propulsion assembly;

[0026] The hydrocarbon substance in-situ detection module includes:

[0027] A photoionization sensor, a non-dispersive infrared sensor, and a fiber-optic near-infrared spectral module, and the hydrocarbon substance in-situ detection module is arranged in a shared holder mechanism of the infrared thermal imager and the rotatable detection holder at the front end of the robot.

[0028] Preferably, the control and communication system includes a core controller and a communication unit.

[0029] The core controller includes a Raspberry Pi 4B main control board and a motion control board.

[0030] The communication unit includes a LoRa remote communication module, a waterproof float, and a waterproof connector.

[0031] The Raspberry Pi 4B main control board exchanges data with the motion control board through a PCIe interface, and the

[0032] The LoRa remote communication module is arranged inside the waterproof float, the LoRa remote communication module is connected with the main control system through a waterproof connector via an RS485 bus, and the waterproof float is physically connected with the robot body through a detachable cable.

[0033] Preferably, the power system further includes a land power system and an underwater propulsion system.

[0034] The land power system includes the following steps:

[0035] Step 1: adopt wheel drive mode, combined with joint motion to realize flexible steering and crawling;

[0036] Step 2: realize serpentine motion, according to the curvature equation of the snake curve simulating the serpentine motion of the biological snake proposed by Professor Hirose through observing the motion law of the biological snake

[0037] In the formula, s represents the arc length of a point on the snake curve to the starting point, L represents the total length of the snake robot body, represents the number of waves contained in the snake robot body, n represents the nth joint, represents the initial angle of the snake curve, and by controlling the motion of each joint, the robot can perform serpentine motion on the ground, realizing efficient land travel;

[0038] Step 3: give the ability to climb stairs, divide the robot into upper plane joint, lower plane joint and connecting joint three parts, detect the step height through laser ranging sensor, judge whether it needs to climb the stairs; if the step height is within the maximum height that the robot can climb, then perform the action of climbing the stairs; otherwise, send a signal to the operator that the step is too high, and wait for other instructions; when climbing the stairs, the head joint of the robot is lifted upward to climb the stairs, and the connecting joints are moved from front to back in turn; when descending the stairs, the robot descends the stairs when the head joint torque is detected to be negative;

[0039] The underwater propulsion system comprises the following steps:

[0040] Step 1: select the propulsion mode, use propeller or bionic fin structure as underwater propulsion mode; select appropriate propulsion structure according to different water environment to adapt to different underwater rescue tasks;

[0041] We select the serpentine motion gait as the motion gait of the bionic water snake robot, and the serpentine motion is also called sinusoidal motion. This motion mode is the most common and fastest type of snake motion. To realize the general sinusoidal motion mode, we need to make each connecting joint i N∈-{1,…,1}

[0042] Satisfy the sinusoidal motion tracking reference signal

[0043] In the formula, is the gait amplitude, ω is the angular frequency, δ is the phase difference between adjacent joints, γ is the constant offset causing turning motion, and g(i,N) function is the scaling function of joint i amplitude. When g(i,N)=i, the motion mode is serpentine motion, and at this time the sinusoidal motion is the serpentine motion connecting joint angle input signal;

[0044] Step two: provide power support, the power source uses high-performance lithium battery pack, equipped with intelligent battery management system, to ensure the safe and stable operation of the battery and efficient energy utilization, to provide power support for the long-term operation of the robot underwater.

[0045] Preferably, the sensor system further comprises a sensor data fusion processing method, the sensor data fusion processing method comprising the following steps:

[0046] Step one: data acquisition, the computer is connected with high-definition camera, infrared thermal imager, multi-gas detector, laser radar, bioelectric field sensor and nine-axis IMU, etc. sensors, real-time acquisition of data from different sensors; the above data contains visual information, temperature information, gas concentration information, distance information, bioelectric signal and the attitude information of the robot, etc.

[0047] Step two: data preprocessing, the computer preprocesses the collected sensor data; first, filter algorithm is used to remove high-frequency noise in the data, such as Gaussian filter, median filter, etc.; then, the data is denoised to further improve the quality of the data; finally, the data is normalized to unify the data of different dimensions and different ranges to a fixed interval, which is convenient for subsequent data fusion and processing;

[0048] Step three: data fusion, the computer uses Kalman filter or multi-sensor data fusion algorithm to fuse the preprocessed data; Kalman filter recursively estimates the sensor data by establishing the dynamic model and observation model of the system to obtain the optimal fusion result; multi-sensor data fusion algorithm considers the characteristics and weights of each sensor, uses weighted average, Bayesian fusion, etc. method, fuses the data of different sensors, so as to improve the accuracy of environment perception and target recognition, and provides more accurate information for path planning and obstacle avoidance.

[0049] Preferably, the control system further comprises: improved adaptive elite ant colony optimization (IAEACO), improved dynamic window approach (DWA), and fusion algorithm execution;

[0050] The improved adaptive elite ant colony optimization (IAEACO) comprises the following steps:

[0051] Step one: initialize the ant colony parameters, the computer sets the number of ants, pheromone intensity, heuristic factor weight and other ant colony parameters to make initial preparation for the operation of the algorithm;

[0052] Step two: Introduce distance parameter factor to improve heuristic function. The computer adds distance parameter factor to the traditional heuristic function, making the ants more inclined to choose the direction with shorter distance when selecting the path, thereby enhancing the global search ability of the algorithm and avoiding falling into local optimum;

[0053] Step three: Adopt pseudo-random state transition rule and adaptive update mechanism. The computer calculates the probability of the ant moving to the next node according to the pseudo-random state transition rule, combining the current position and state of the ant. At the same time, according to the iteration situation of the algorithm and the search effect of the ant, the update strength and mode of the pheromone are adaptively adjusted to balance the global search ability and local search ability of the algorithm, and to optimize the path selection process;

[0054] Step four: Integrate angle guiding factor into transition probability function. The computer integrates the angle guiding factor into the transition probability function, so that the ant not only considers the distance factor when selecting the path, but also considers the directionality of the path. This can guide the ant to search more efficiently in the target direction, enhance the purpose of the algorithm, and speed up the convergence speed;

[0055] Step five: Reward strategy based on elite ant model. In each iteration, the computer selects the elite ants with excellent performance and rewards the paths they have passed through, increasing the pheromone concentration of these paths. This can strengthen the guiding effect of the shortest path on subsequent iterations, making the algorithm converge to the global optimal solution more quickly;

[0056] The improved dynamic window approach (DWA) includes the following steps:

[0057] Step one: Add global path evaluation function. The computer integrates the global path evaluation function into the evaluation function of the traditional DWA algorithm. This evaluation function considers multiple factors such as path distance, smoothness, safety, etc., so that the algorithm can fully refer to the information of the global path when planning the local path, avoid falling into local optimum, and improve the globality and reliability of navigation;

[0058] Step two: Combine IAEACO algorithm to optimize path planning. The computer uses the global optimal path generated by the IAEACO algorithm as the reference path for the DWA algorithm. Based on this, the DWA algorithm adjusts the speed and direction of the robot in real time according to the current dynamic environment information of the robot, plans the local path and avoids obstacles, thereby improving the navigation accuracy and real-time performance;

[0059] The fusion algorithm execution includes the following steps:

[0060] Step one: generate the global optimal path: the computer searches and optimizes the map according to the environmental data collected by sensors such as lidar and high-definition camera, and generates a global optimal path from the starting point to the target point;

[0061] Step two: real-time obstacle avoidance and dynamic path adjustment: the computer sends the generated global optimal path to the DWA algorithm module; the DWA algorithm adjusts the robot's motion speed and direction in real time according to the robot's current environmental perception information such as the distance to the obstacle, the motion state of the obstacle, etc., and performs local path planning and obstacle avoidance to ensure the robot's safe and efficient movement along the global optimal path in a dynamic environment;

[0062] Step three: output control instructions, the computer outputs corresponding control instructions according to the real-time path planning results generated by the DWA algorithm, drives the robot's motor, rudder and other actuators, and makes the robot move according to the planned path, realizing autonomous exploration.

[0063] Preferably, the control and communication system further comprises a path planning and obstacle avoidance method; the path planning and obstacle avoidance method comprises the following steps:

[0064] Step one: map construction, the computer constructs a robot working map according to the environmental information collected by lidar and high-definition camera using simultaneous localization and mapping (SLAM) technology; the map contains information such as the layout of the environment, the position and shape of the obstacles, etc., providing a basis for path planning;

[0065] Step two: global optimal path planning, the computer uses the IAEACO algorithm to search and optimize the path based on the constructed map, taking the current position of the robot as the starting point and the target point as the end point; the IAEACO algorithm simulates the foraging behavior of ants, constantly updates the pheromone concentration of the path, guides ants to find the optimal path, and finally calculates a global optimal path from the starting point to the target point;

[0066] Step three: real-time obstacle avoidance and dynamic path adjustment, the computer obtains obstacle information detected by lidar and high-definition camera in real time during the robot's movement; combined with the DWA algorithm, the robot's motion speed and direction are dynamically adjusted according to the position, speed and direction of the obstacle, local path planning and obstacle avoidance are performed; the DWA algorithm quickly searches for feasible motion trajectories in the dynamic window at the robot's current position, and selects the optimal trajectory for motion, ensuring the robot's safe and efficient movement in a complex dynamic environment;

[0067] Step four: robot motion control, the computer outputs corresponding control instructions according to the real-time adjusted path planning result, drives the chassis, steering engine and other actuators of the robot, controls the robot to move along the optimized path, and realizes autonomous exploration.

[0068] Compared with the prior art, the beneficial effects of the present application are:

[0069] (1) By adding soil, groundwater and hydrocarbon in-situ sensor systems, the "rescue-monitoring" integrated capability is formed, the environmental detection dimension is expanded, the "rescue + risk assessment" dual-mode cooperation is realized, and the detection reliability and data representativeness in complex environments are also improved.

[0070] (2) The snake-like multi-degree-of-freedom motion realized by the modular flexible joint structure, combined with the cooperative control of wheeled driving and underwater propulsion and the built-in improved IAEACO algorithm, can make the robot have excellent terrain adaptability, can easily enter complex terrains such as ruins, caves and shoals, break through the limitations of traditional rescue equipment, and open up new channels for rescue. Its motion flexibility is improved by more than 60% compared with traditional wheeled / caterpillar robots, and it can more freely avoid obstacles, adjust posture and efficiently perform rescue tasks in complex environments.

[0071] (3) Through the cooperative work of multi-modal sensors, environmental information is obtained from different dimensions, and multi-source data fusion is realized through Kalman filtering to improve data accuracy and reliability. Subsequently, the LoRa module (403) is used to ensure real-time data transmission, ensuring that rescue personnel can obtain real-time information. The robot is equipped with an advanced environmental perception system, and the environmental perception response time is less than 200 milliseconds. The target recognition accuracy is as high as 95% or more, and it can accurately identify targets such as trapped personnel. It can also detect more than 6 types of dangerous gases, with a detection sensitivity of ppm level, and can provide early warning of potential dangers.

[0072] (4) Through the design of the propulsion assembly and the waterproof universal coupling, combined with the smooth transition of the motion mode realized by the fusion of the DWA algorithm, the robot can quickly switch between water and land environments, with a mode switching time of less than 5 seconds and a maximum underwater speed of 2 knots. Compared with existing amphibious robots, it reduces weight by 30% and energy consumption by 25%, prolongs endurance time, and expands the rescue range. BRIEF DESCRIPTION OF DRAWINGS

[0073] Figure 1 is a mechanical structure perspective view of the device of the present application;

[0074] Figure 2 is one of the sensor distribution perspective views of the present application;

[0075] Figure 3Figure 2 is a perspective view of a sensor distribution of the present application;

[0076] Figure 4 Figure 3 is a perspective view of a control system architecture of the present application;

[0077] Figure 5 Figure 4 is a flowchart of an adaptive elite ant colony optimization (IAEACO) of the present application;

[0078] Figure 6 Figure 5 is a flowchart of a path planning and obstacle avoidance method of the present application;

[0079] In the figure: 101, main trunk unit; 102, flexible connection unit; 103, wheeled drive assembly; 104, propulsion assembly; 201, lithium battery pack; 202, intelligent battery management system; 203, wheeled drive motor; 204, reduction gear set; 205, underwater propulsion motor; 206, universal coupling; 301, high-definition camera; 302, infrared thermal imager; 303, laser radar; 304, nine-axis IMU; 305, depth sensor; 306, multi-gas detector; 307, bioelectric field sensor; 401, Raspberry Pi 4B main control board; 402, motion control board; 403, LoRa remote communication module; 404, waterproof buoy; 405, waterproof connector; 501, soil moisture / conductivity sensor; 502, soil pH sensor; 503, soil heavy metal ion sensor; 601, dissolved oxygen (DO) sensor; 602, turbidity sensor; 603, total organic carbon (TOC) sensor; 604, heavy metal rapid detector; 701, photoionization (PID) sensor; 702, non-dispersive infrared (NDIR) sensor; 703, fiber optic near-infrared spectroscopy module. DETAILED DESCRIPTION

[0080] The technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all embodiments. Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art without creative labor fall within the scope of the present application.

[0081] In the description of the present application, it should be understood that the terms "center", "longitudinal", "transverse", "upper", "lower", "front", "back", "left", "right", "vertical", "horizontal", "top", "bottom", "inner", "outer", and the like indicate the orientation or positional relationship based on the orientation or positional relationship shown in the drawings, and are only for the convenience of describing the present application and simplifying the description, and do not indicate or imply that the device or element referred to must have a particular orientation, be constructed and operated in a particular orientation, and therefore cannot be understood as a limitation on the present application. In addition, the terms "first", "second", and the like are only for descriptive purposes and cannot be understood as indicating or implying relative importance or implicitly indicating the number of technical features indicated. Therefore, the features defined with "first", "second", and the like can explicitly or implicitly include one or more of the features. In the description of the present application, unless otherwise stated, the meaning of "a plurality of" is two or more.

[0082] Embodiment one:

[0083] Please refer to Figures 1-2 As shown in the figure, an amphibious snake robot, comprising:

[0084] Mechanical structure, power system, sensor system, control and communication system;

[0085] The mechanical structure comprises:

[0086] The main trunk unit 101 is made of a rigid segment of high-strength lightweight alloy;

[0087] The flexible connection unit 102 is a bendable joint made of a tough composite material, used to connect adjacent main trunk units 101, and the main trunk unit 101 and the flexible connection unit 102 are connected in series to form a snake-like bionic structure, providing basic support and power load;

[0088] The motion execution mechanism includes a wheeled drive assembly 103 and a propulsion assembly 104, the wheeled drive assembly 103 is installed at the bottom of the main trunk unit 101, the wheeled drive assembly 103 is rigidly connected with the main trunk unit 101 to realize the function of land movement, the propulsion assembly 104 adopts a switchable propeller and a bionic fin assembly, and the propulsion assembly 104 is arranged at the end of the trunk unit, the propulsion assembly 104 is connected with the end main trunk unit 101 through a quick connector, realizing the switching of underwater propulsion mode;

[0089] From the above, the bendable joint and multi-mode drive: first in the bionic mechanical structure, the modular bendable joint design (flexible connection unit 102) is combined with the rigid segment of the main trunk unit 101, and a bionic snake-shaped multi-degree-of-freedom motion mechanism with more than 8 degrees of freedom is constructed. Compared with the traditional rigid structure robot, the minimum bending radius is reduced from ≥30cm to 15cm, the narrow space passing capacity is improved by 120%, and the terrain adaptability is improved from ≤60% to 92%. Subsequently, in the amphibious driving system link, through the magnetic attraction type quick connection mechanism (switching time <5s), the wheel driving assembly 103 and the propeller / bionic fin of the propulsion assembly 104 are switched in dual mode, so that the water-land mode switching efficiency is improved by 600% compared with the traditional 30 seconds or more. At the same time, the underwater endurance of 4 hours (traditional ≤2 hours) and the maximum climbing ability of 45° (traditional ≤30°) are achieved, and the performance breakthrough is achieved. Finally, a complete technical solution is formed that the shape is adaptive and the multi-mode drive is coordinated.

[0090] From Figure 1 , Figure 3 and Figure 4 it can be seen that the power system comprises:

[0091] an energy module and a power transmission system;

[0092] The energy module comprises: a lithium battery pack 201 and an intelligent battery management system 202, the lithium battery pack 201 is distributed along the axis of the main trunk unit 101, and the intelligent battery management system 202 is electrically connected with each battery pack;

[0093] The power transmission system comprises: a wheel driving motor 203 and an underwater propulsion motor 205, the wheel driving motor 203 is connected with the driving wheel through a reduction gear set 204, and the underwater propulsion motor 205 is connected with the propeller and bionic fin assembly of the propulsion assembly 104 through a universal joint 206.

[0094] Specifically, regarding the above-mentioned sensor system, as shown in Figure 2 , the sensor system comprises:

[0095] an environment perception module, a pose detection module, a special sensor module, a soil parameter in-situ detection module, a groundwater quality in-situ monitoring module, and a hydrocarbon in-situ detection module;

[0096] The environment perception module comprises: a high-definition camera 301, an infrared thermal imager 302, and a laser radar 303, the high-definition camera 301 and the infrared thermal imager 302 form a binocular vision unit, and the laser radar 303 is installed at the front end of the robot;

[0097] The pose detection module comprises a nine-axis IMU 304 and a depth sensor 305, the nine-axis IMU 304 is internally provided with a gyroscope, an accelerometer and a magnetometer, and the depth sensor 305 is arranged at the bottom of the robot;

[0098] The special sensor comprises a multi-gas detector 306 and a bioelectric field sensor 307;

[0099] The high-definition camera 301 and the infrared thermal imager 302 are connected with the first main trunk unit 101 through a cloud platform mechanism, three-dimensional reconstruction of the environment is realized, the laser radar 303 is directly connected with the main control system, a SLAM environment map is established, the nine-axis IMU 304 is fixed at the center of mass of the robot, and pose data is transmitted through an I2C interface;

[0100] Specifically, regarding the above-mentioned soil parameter in-situ detection module, refer to Figure 3 The soil parameter in-situ detection module comprises:

[0101] The soil humidity / conductivity sensor 501, the soil pH sensor 502 and the soil heavy metal ion sensor 503; the soil parameter in-situ detection module is arranged at the bottom of the middle main trunk unit 101 of the robot, is designed to retract the spring-driven telescopic probe to be flush with the trunk, is extended by 5 cm to contact the soil in the land mode, is automatically locked and retracted under the water mode, and has a protection level of IP68;

[0102] The underground water quality in-situ monitoring module comprises:

[0103] The dissolved oxygen DO sensor 601, the turbidity sensor 602, the total organic carbon TOC sensor 603 and the heavy metal rapid detector 604, and the underground water quality in-situ monitoring module is integrated in the waterproof detection cabin outside the terminal propulsion assembly 104;

[0104] The hydrocarbon in-situ detection module comprises:

[0105] The photoionization PID sensor 701, the non-dispersive infrared NDIR sensor 702 and the optical fiber near-infrared spectrum module 703, and the hydrocarbon in-situ detection module is arranged in the rotatable detection cloud platform at the front end of the robot and shares the cloud platform mechanism with the high-definition camera 301 and the infrared thermal imager 302;

[0106] From the above, when traveling on land, the probe is inserted into the surface soil (0-5 cm deep) after the terrain stability is judged by the pose detection module (304), and 1 set of data is collected per second; in complex terrain (such as rubble gaps), the exposed soil area is located by the laser radar 303, and the trunk posture is actively adjusted to complete the fixed-point detection. After the data is preprocessed (filtering, temperature compensation), it is transmitted to the Raspberry Pi 4B main control board 401 through the I2C interface; when in underwater mode, the depth sensor 305 detects that the water depth is greater than 0.5 m, and then automatically activates, the sensor cabin introduces water flow through the flow guide groove, and combines with the low-speed rotation of the propulsion assembly 104 to realize water circulation contact; for turbid water, use the infrared thermal imager 302 to compensate for light attenuation, improve the stability of TOC and turbidity detection, and transmit the data to the intelligent battery management system 202 through the waterproof cable through the CAN bus. The underwater sonar data is fused to construct the water quality distribution profile; when in land mode, the gimbal is expanded to the horizontal position, and the three-dimensional laser radar 303 is used to locate the suspected leakage point (such as the damaged part of the pipeline), and the PID / NDIR sensor actively inhales sampling (with a built-in micro-pump, sampling flow rate 1L / min); when in underwater mode, the gimbal is contracted to a streamlined posture, and the optical fiber module directly contacts the water body. Use the original SLAM algorithm (±2cm accuracy) to mark the coordinates of the abnormal area of hydrocarbon concentration.

[0107] Referring to Figure 4 , the control and communication system includes a core controller and a communication unit;

[0108] The core controller includes a Raspberry Pi 4B main control board 401 and a motion control board 402;

[0109] The communication unit includes a LoRa remote communication module 403, a waterproof buoy 404, and a waterproof connector 405;

[0110] The Raspberry Pi 4B main control board 401 interacts with the motion control board 402 through the PCIe interface, the LoRa remote communication module 403 is placed inside the waterproof buoy 404, the LoRa remote communication module 403 is connected with the main control system through the RS485 bus through the waterproof connector 405, and the waterproof buoy 404 is physically connected with the robot body through a detachable cable;

[0111] Embodiment two:

[0112] Referring to Figure 5 , Figure 6 , the power system further includes a land power system and an underwater propulsion system;

[0113] The land power system includes the following steps:

[0114] Step 1: adopt a wheel drive mode, and realize flexible turning and crawling in combination with joint motion;

[0115] Step 2: Implement serpentine motion, according to the curvature equation of the snake curve that simulates the serpentine motion of biological snakes proposed by Professor Hirose through observation of the motion law of biological snakes

[0116] In the formula, s represents the arc length between a point on the snake curve and the starting point, L represents the total length of the snake robot body, represents the number of waves contained in the snake robot body, n represents the nth joint, represents the initial angle of the snake curve, and by controlling the motion of each joint, the robot can perform serpentine motion on the ground, realizing efficient land travel;

[0117] Step 3: Give the ability to climb stairs, divide the robot into three parts: upper plane joint, lower plane joint and connecting joint, detect the height of the stairs through laser ranging sensor, judge whether it needs to climb the stairs; if the height of the stairs is within the maximum height that the robot can climb, then perform the action of climbing the stairs; otherwise, send a signal to the operator that the stairs are too high, and wait for other instructions; when climbing the stairs, the head joint of the robot is lifted upwards to climb the stairs, and the connecting joints are moved from front to back in turn; when descending the stairs, the head joint torque is detected to be negative when descending the stairs;

[0118] The underwater propulsion system includes the following steps:

[0119] Step 1: Choose the propulsion method, use propeller or bionic fin structure as underwater propulsion method; choose appropriate propulsion structure according to different water environment to adapt to different underwater rescue tasks;

[0120] We choose the serpentine motion gait as the motion gait of the bionic water snake robot, and the serpentine motion is also called sinusoidal motion, which is the most common and fastest type of snake motion; to realize the general sinusoidal motion mode, we need to make each connecting joint i N∈-{1,…,1}

[0121] Satisfy the sinusoidal motion tracking reference signal

[0122] In the formula, is the gait amplitude, ω is the angular frequency, δ is the phase difference between adjacent joints, γ is the constant offset that causes turning motion, and g(i,N) function is the scaling function of joint i amplitude, when g(i,N) = i, the motion mode is serpentine motion, at this time the sinusoidal motion is the serpentine motion connecting joint angle input signal;

[0123] Step 2: Provide power support, the power source uses high-performance lithium battery pack 201, equipped with intelligent battery management system 202, to ensure the safe and stable operation of the battery and efficient energy utilization, and to provide power support for the long-time operation of the robot underwater.

[0124] Example three:

[0125] Reference Figure 5 , and Figure 6 As shown in FIG. 1, the sensor system further comprises a sensor data fusion processing method, which comprises the following steps:

[0126] Step 1: Data acquisition. The computer collects data from different sensors in real time by connecting with high-definition camera 301, infrared thermal imager 302, multi-gas detector 306, laser radar 303, bioelectric field sensor 307, nine-axis IMU 304, etc. The above data contains visual information, temperature information, gas concentration information, distance information, bioelectric signals, and robot posture information, etc.

[0127] Step 2: Data preprocessing. The computer performs preprocessing operation on the collected sensor data. First, filter algorithm is used to remove high-frequency noise in the data, such as Gaussian filter, median filter, etc. Then, the data is denoised to further improve the quality of the data. Finally, the data is normalized to unify different dimensions and ranges of data into a fixed interval, which is convenient for subsequent data fusion and processing.

[0128] Step 3: Data fusion. The computer uses Kalman filter or multi-sensor data fusion algorithm to fuse the preprocessed data. Kalman filter establishes dynamic model and observation model of the system to recursively estimate the sensor data and obtain the optimal fusion result. Multi-sensor data fusion algorithm considers the characteristics and weights of each sensor, uses weighted average, Bayesian fusion, etc. to fuse the data of different sensors, thereby improving the accuracy of environment perception and target recognition, and providing more accurate information for path planning and obstacle avoidance.

[0129] From the above, the multi-source data fusion architecture: first through the heterogeneous sensor fusion of laser radar 303 and binocular vision, combined with Kalman filter optimization SLAM accuracy, while integrating bioelectric field sensor 307 and multi-gas detector 306, expand the target detection dimension; then use nine-axis IMU 304 and visual SLAM fusion positioning, and introduce laser radar and infrared thermal imaging 302 cooperative calibration, so that the environmental modeling accuracy is improved to ±2cm (traditional ±5cm), the target recognition accuracy is 98.7% (traditional ≤85%), and the multi-source data fusion efficiency is improved by 300%. Redundant design of water-land communication: through the separation deployment of LoRa remote communication module 403 and waterproof buoy 405, combined with RS485 bus to ensure the stability of underwater signal, avoid the signal attenuation problem of traditional wireless transmission, the communication distance is expanded to 5km (traditional ≤1km). At the same time, based on the hierarchical control architecture, the system integration is optimized, the control response time is less than 50ms, the system reliability MTBF is greater than or equal to 5000 hours, forming an intelligent solution of high-precision perception and stable communication cooperation.

[0130] The control system further comprises: an improved adaptive elite ant colony optimization algorithm (IAEACO), an improved dynamic window approach (DWA), and a fusion algorithm execution;

[0131] The improved adaptive elite ant colony optimization algorithm (IAEACO) comprises the following steps:

[0132] Step 1: initialize the ant colony parameters, set the number of ants, pheromone intensity, heuristic factor weight and other ant colony parameters on the computer, and make initial preparations for the operation of the algorithm;

[0133] Step 2: introduce a distance parameter factor to improve the heuristic function, add a distance parameter factor to the traditional heuristic function on the computer, so that the ant is more inclined to choose a shorter distance direction when selecting a path, thereby enhancing the global search ability of the algorithm and avoiding falling into a local optimum;

[0134] Step 3: adopt a pseudo-random state transition rule and an adaptive update mechanism, calculate the probability of the ant moving to the next node according to the pseudo-random state transition rule and the current position and state of the ant; at the same time, according to the iteration of the algorithm and the search effect of the ant, the update strength and mode of the pheromone are adaptively adjusted to balance the global search ability and the local search ability of the algorithm, and the path selection process is optimized;

[0135] Step four: fuse the angle guidance factor into the transition probability function, the computer integrates the angle guidance factor into the transition probability function, so that the ants consider not only the distance factor but also the directionality of the path when selecting the path; this can guide the ants to search more efficiently towards the target direction, enhance the purpose of the algorithm, and speed up the convergence speed;

[0136] Step five: elite ant model-based reward strategy, the computer selects the elite ants with excellent performance in each iteration and rewards the paths they have passed through, increasing the pheromone concentration of these paths; this can strengthen the guiding effect of the shortest path on subsequent iterations, making the algorithm converge to the global optimal solution more quickly;

[0137] The improved dynamic window method DWA includes the following steps:

[0138] Step one: add a global path evaluation function, the computer integrates a global path evaluation function into the evaluation function of the traditional DWA algorithm; this evaluation function considers multiple factors such as path distance, smoothness, safety, etc., so that the algorithm can fully refer to the information of the global path when performing local path planning, avoid falling into local optimization, and improve the globality and reliability of navigation;

[0139] Step two: combine IAEACO algorithm to optimize path planning: the computer uses the global optimal path generated by the IAEACO algorithm as the reference path for the DWA algorithm; based on this, the DWA algorithm adjusts the speed and direction of the robot in real time according to the current dynamic environmental information of the robot, performs local path planning and obstacle avoidance, thereby improving navigation accuracy and real-time performance;

[0140] The fusion algorithm execution includes the following steps:

[0141] Step one: generate a global optimal path: the computer uses the IAEACO algorithm to search and optimize the map based on environmental data collected by sensors such as laser radar 303 and high-definition camera 301, generating a global optimal path from the starting point to the target point;

[0142] Step two: real-time obstacle avoidance and dynamic path adjustment: the computer sends the generated global optimal path to the DWA algorithm module; the DWA algorithm adjusts the motion speed and direction of the robot in real time based on the current environmental perception information of the robot, such as the distance to the obstacle and the motion state of the obstacle, performs local path planning and obstacle avoidance, and ensures that the robot moves safely and efficiently along the global optimal path in a dynamic environment;

[0143] Step three: output control instructions. The computer outputs corresponding control instructions according to the real-time path planning result generated by the DWA algorithm, drives the motor, steering gear and other actuators of the robot, and makes the robot move according to the planned path, thereby realizing autonomous exploration.

[0144] The control and communication system further comprises a path planning and obstacle avoidance method, which comprises the following steps:

[0145] Step one: map construction. The computer constructs a working map of the robot based on the environmental information collected by the laser radar 303 and the high-definition camera 301 using the simultaneous localization and mapping (SLAM) technology. The map contains information such as the layout of the environment, the position and shape of obstacles, etc., providing a basis for path planning.

[0146] Step two: global optimal path planning. Based on the constructed map, the computer takes the current position of the robot as the starting point and the target point as the end point, and uses the IAEACO algorithm to search and optimize the path. The IAEACO algorithm simulates the foraging behavior of ants, constantly updates the pheromone concentration of the path, and guides the ants to find the optimal path. Finally, a global optimal path from the starting point to the target point is calculated.

[0147] Step three: real-time obstacle avoidance and dynamic path adjustment. During the movement of the robot, the computer obtains real-time obstacle information detected by the laser radar 303 and the high-definition camera 301. Combined with the DWA algorithm, the computer dynamically adjusts the movement speed and direction of the robot according to the position, speed and direction of the obstacles, and performs local path planning and obstacle avoidance. The DWA algorithm quickly searches for feasible movement trajectories within the dynamic window at the current position of the robot, and selects the optimal trajectory for movement, ensuring that the robot moves safely and efficiently in complex dynamic environments.

[0148] Step four: robot motion control. The computer outputs corresponding control instructions according to the real-time adjusted path planning result, drives the chassis, steering gear and other actuators of the robot, and controls the robot to move along the optimized path, thereby realizing autonomous exploration.

[0149] IAEACO-DWA hybrid algorithm: First, the angle guiding factor and elite strategy are introduced into the improved ant colony algorithm (IAEACO) to optimize the global path search efficiency. Then, the dynamic window algorithm (DWA) is embedded into the global evaluation function to realize real-time collaboration of dynamic obstacle avoidance and global path, effectively reducing trajectory oscillation. Compared with traditional path planning algorithms (such as A*), this hybrid algorithm combines a dynamic environment adaptive mechanism, shortens the path planning time from ≥5s to 0.8s, improves the obstacle avoidance success rate from ≤85% to 99%, and increases the complex environment pass rate by 150%, significantly enhancing the autonomous decision-making ability of the robot in dynamic scenarios.

[0150] While embodiments of the present application have been shown and described, it is to be understood that the embodiments described are merely divergences and modifications and not limitations of the scope of the application, which is defined by the appended claims and their equivalents.

Claims

1. An amphibious snake-like robot, characterized in that, include: Mechanical structure, power system, sensor system, control and communication system; The mechanical structure includes: The main body unit (101) is a rigid segment made of a high-strength, lightweight alloy; The flexible connection unit (102), which is a bendable joint made of tough composite material, is used to connect adjacent main body units (101). The main body units (101) and the flexible connection unit (102) are connected in series to form a snake-like bionic structure, providing basic support and dynamic load bearing. The motion actuator includes a wheel drive assembly (103) and a propulsion assembly (104). The wheel drive assembly (103) is installed at the bottom of the main body unit (101) and is rigidly connected to the main body unit (101) to realize land movement. The propulsion assembly (104) adopts a switchable propeller and a biomimetic fin assembly and is arranged at the end body unit. The propulsion assembly (104) is connected to the end main body unit (101) through a quick interface to realize underwater propulsion mode switching.

2. An amphibious snake-like robot according to claim 1, characterized in that: The power system includes: Energy modules and power transmission systems; The energy module includes: a lithium battery pack (201) and an intelligent battery management system (202). The lithium battery pack (201) is distributed along the axial direction of the main body unit (101), and the intelligent battery management system (202) is electrically connected to each battery pack. The power transmission system includes a wheel drive motor (203) and an underwater propulsion motor (205). The wheel drive motor (203) is connected to the drive wheel through a reduction gear set (204), and the underwater propulsion motor (205) is connected to the propeller and bionic fin assembly of the propulsion assembly (104) through a universal coupling (206).

3. An amphibious snake-like robot according to claim 1, characterized in that: The sensor system includes: The system includes an environmental perception module, a pose detection module, a special sensing module, an in-situ soil parameter detection module, an in-situ groundwater quality monitoring module, and an in-situ hydrocarbon detection module. The environmental perception module includes: a high-definition camera (301), an infrared thermal imager (302), and a lidar (303). The high-definition camera (301) and the infrared thermal imager (302) form a binocular vision unit, and the lidar (303) is installed at the head of the robot. The pose detection module includes a nine-axis IMU (304) and a depth sensor (305). The nine-axis IMU (304) has a built-in gyroscope, accelerometer and magnetometer. The depth sensor (305) is arranged on the bottom of the robot. The special sensors include: a multi-gas detector (306) and a bioelectric field sensor (307); The high-definition camera (301) and the infrared thermal imager (302) are connected to the first main body unit (101) through a gimbal mechanism to realize three-dimensional reconstruction of the environment. The lidar (303) is directly connected to the main control system to establish a SLAM environment map. The nine-axis IMU (304) is fixed at the center of mass of the robot and transmits pose data through the I2C interface.

4. An amphibious snake-like robot according to claim 3, characterized in that: The in-situ soil parameter detection module includes: Soil moisture / conductivity sensor (501), soil pH sensor (502), and soil heavy metal ion sensor (503); the soil parameter in-situ detection module is set at the bottom of the robot's main body unit (101) in the middle section, and is designed with a spring-driven telescopic probe (which is flush with the body when retracted, extends 5cm to contact the soil in land mode, and automatically locks and retracts in underwater mode), with a protection level of IP68. The in-situ groundwater quality monitoring module includes: Dissolved oxygen (DO) sensor (601), turbidity sensor (602), total organic carbon (TOC) sensor (603) and heavy metal rapid detector (604), the groundwater quality in-situ monitoring module is integrated in the waterproof detection chamber outside the end propulsion assembly (104); The in-situ hydrocarbon detection module includes: The system includes a photoionization (PID) sensor (701), a nondispersive infrared (NDIR) sensor (702), and a fiber optic near-infrared spectroscopy module (703). The in-situ hydrocarbon detection module is deployed on a rotatable detection gimbal at the front of the robot (sharing a gimbal mechanism with the high-definition camera (301) and the infrared thermal imager (302)).

5. An amphibious snake-like robot according to claim 1, characterized in that: The control and communication system includes: a core controller and a communication unit; The core controller includes: a Raspberry Pi 4B main control board (401) and a motion control board (402); The communication unit includes: a LoRa remote communication module (403), a waterproof buoy (404), and a waterproof connector (405); The Raspberry Pi 4B main control board (401) interacts with the motion control board (402) via the PCIe interface. The LoRa remote communication module (403) is placed inside the waterproof buoy (404). The LoRa remote communication module (403) is connected to the main control system via the RS485 bus and the waterproof connector (405). The waterproof buoy (404) is physically connected to the robot body via a detachable cable.

6. An amphibious snake-like robot according to claim 1, characterized in that: The power system also includes: a land-based power system and an underwater propulsion system; The land-based power system includes the following steps: Step 1: Employ a wheeled drive system combined with joint movements to achieve flexible steering and crawling; Step 2: Achieve meandering motion, based on the curvature equation of the serpentine curve that Professor Hirose proposed by observing the movement patterns of snakes to mimic their meandering motion. In the formula, s represents the arc length between a point on the snake curve and the starting point, L represents the total length of the snake robot's body, represents the number of waveforms contained in the snake robot's body, n represents the nth joint, and represents the initial angle of the snake curve. By controlling the movement of each joint, the robot can perform meandering movements on the ground, achieving efficient land travel. Step 3: Assign the robot the ability to climb stairs. Divide the robot into three parts: upper plane joints, lower plane joints, and connecting joints. Use a laser rangefinder to detect the height of the stairs and determine if climbing is necessary. If the height is within the robot's maximum climbing height, it will perform the climbing action; otherwise, it will send a signal to the operator that the stairs are too high and wait for other instructions. When climbing stairs, the robot's head joint lifts up to climb the stairs, and the connecting joints move sequentially from front to back. When descending stairs, the robot descends when a negative torque is detected in the head joint. The underwater propulsion system includes the following steps: Step 1: Select the propulsion method, using a propeller or biomimetic fin structure as the underwater propulsion method; select the appropriate propulsion structure according to different water environments to adapt to different underwater rescue missions; We selected a serpentine gait as the movement pattern for the biomimetic water snake robot. Serpentine motion, also known as sinusoidal motion, is the most common and fastest type of snake movement. To achieve a general sinusoidal motion pattern, we need to ensure that each joint i of the water snake robot is N∈-{1,…,1} Satisfying the sinusoidal motion tracking reference signal In the formula, is the gait amplitude, ω is the angular frequency, δ is the phase difference between adjacent joints, γ is the constant offset that causes the turning motion, and g(i,N) is the scaling function of the amplitude of joint i. When g(i,N)=i, the motion mode is serpentine motion. At this time, the sinusoidal motion, i.e., serpentine motion, is connected to the joint angle input signal. Step 2: Provide power support. The power source adopts a high-performance lithium battery pack (201) and is equipped with an intelligent battery management system (202) to ensure the safe and stable operation of the battery and efficient energy utilization, providing power support for the robot to operate underwater for a long time.

7. An amphibious snake-like robot according to claim 1, characterized in that: The sensor system also includes a sensor data fusion processing method, which includes the following steps: Step 1: Data Acquisition. The computer is connected to sensors such as a high-definition camera (301), an infrared thermal imager (302), a multi-gas detector (306), a lidar (303), a bioelectric field sensor (307), and a nine-axis IMU (304) to collect data from different sensors in real time. The data includes visual information of the environment, temperature information, gas concentration information, distance information, bioelectric signals, and robot posture information. Step 2: Data preprocessing. The computer performs preprocessing operations on the collected sensor data. First, filtering algorithms are used to remove high-frequency noise from the data, such as Gaussian filtering and median filtering. Then, the data is denoised to further improve the data quality. Finally, the data is normalized to unify data of different dimensions and ranges into a fixed interval, which facilitates subsequent data fusion and processing. Step 3: Data Fusion. The computer uses Kalman Filter or multi-sensor data fusion algorithms to fuse the preprocessed data. Kalman Filter establishes a dynamic model and an observation model of the system to recursively estimate the sensor data and obtain the optimal fusion result. Multi-sensor data fusion algorithms comprehensively consider the characteristics and weights of each sensor and use methods such as weighted averaging and Bayesian fusion to fuse data from different sensors, thereby improving the accuracy of environmental perception and target recognition and providing more accurate information for path planning and obstacle avoidance.

8. An amphibious snake-like robot according to claim 1, characterized in that: The control system also includes: an improved adaptive elite ant colony optimization (IAEACO) algorithm, an improved dynamic window approach (DWA) algorithm, and a fusion algorithm execution. The improved adaptive elite ant colony algorithm (IAEACO) includes the following steps: Step 1: Initialize ant colony parameters. The computer sets ant colony parameters such as the number of ants, pheromone intensity, and heuristic factor weight to prepare for the algorithm's operation. Step 2: Introduce a distance parameter factor to improve the heuristic function. Based on the traditional heuristic function, the computer adds a distance parameter factor, making the ants more inclined to choose the shorter path, thereby enhancing the algorithm's global search ability and avoiding getting trapped in local optima. Step 3: Using pseudo-random state transition rules and an adaptive update mechanism, the computer calculates the probability of the ant moving to the next node based on the pseudo-random state transition rules and the current position and state of the ant. At the same time, based on the algorithm's iteration and the ant's search performance, the computer adaptively adjusts the intensity and method of pheromone updates to balance the algorithm's global search capability and local search capability, thereby optimizing the path selection process. Step 4: Integrate the angle guiding factor into the transition probability function. The computer incorporates the angle guiding factor into the transition probability function, so that when the ant chooses a path, it considers not only the distance factor but also the direction of the path. This can guide the ant to search more efficiently in the target direction, enhance the purposefulness of the algorithm, and speed up the convergence speed. Step 5: Based on the reward strategy of the elite ant model, the computer selects the best-performing elite ants in each iteration and rewards the paths they have taken, increasing the pheromone concentration of these paths. This strengthens the guiding role of the shortest path in subsequent iterations, enabling the algorithm to converge to the global optimum more quickly. The improved Dynamic Window Method (DWA) includes the following steps: Step 1: Incorporate a global path evaluation function. The computer integrates a global path evaluation function into the evaluation function of the traditional DWA algorithm. This evaluation function comprehensively considers multiple factors such as path distance, smoothness, and safety, enabling the algorithm to fully refer to global path information when planning local paths, avoiding getting trapped in local optima, and improving the globality and reliability of navigation. Step 2: Optimize path planning using the IAEACO algorithm: The computer uses the globally optimal path generated by the IAEACO algorithm as the reference path for the DWA algorithm; based on this, the DWA algorithm adjusts the robot's speed and direction in real time according to the robot's current dynamic environment information, performs local path planning and obstacle avoidance, thereby improving navigation accuracy and real-time performance. The fusion algorithm is executed by the following steps: Step 1: Generate the globally optimal path: The computer uses the IAEACO algorithm to search and optimize the map based on environmental data collected by sensors, such as data from LiDAR (303) and HD cameras (301), to generate the globally optimal path from the starting point to the target point. Step 2: Real-time obstacle avoidance and dynamic path adjustment: The computer sends the generated global optimal path to the DWA algorithm module; the DWA algorithm adjusts the robot's speed and direction in real time based on the robot's current environmental perception information, such as the distance to obstacles and the movement state of obstacles, to perform local path planning and obstacle avoidance, ensuring that the robot moves safely and efficiently along the global optimal path in a dynamic environment. Step 3: Output control commands. Based on the real-time path planning results generated by the DWA algorithm, the computer outputs corresponding control commands to drive the robot's motors, servos, and other actuators, enabling the robot to move along the planned path and achieve autonomous exploration.

9. An amphibious snake-like robot according to claim 1, characterized in that: The control and communication system further includes: a path planning and obstacle avoidance method; the path planning and obstacle avoidance method includes the following steps: Step 1: Map building. Based on the environmental information collected by the LiDAR (303) and HD camera (301), the computer uses Simultaneous Localization and Mapping (SLAM) technology to build a working map for the robot. The map contains information such as the layout of the environment and the location and shape of obstacles, which provides a basis for path planning. Step 2: Global optimal path planning. Based on the constructed map, the computer uses the IAEACO algorithm to search and optimize the path, starting from the robot's current position and ending at the target point. The IAEACO algorithm continuously updates the pheromone concentration of the path by simulating the foraging behavior of ants, guiding the ants to find the optimal path, and finally calculating a globally optimal path from the starting point to the target point. Step 3: Real-time obstacle avoidance and dynamic path adjustment. During the robot's movement, the computer acquires obstacle information detected by the LiDAR (303) and HD camera (301) in real time. Combined with the DWA algorithm, the robot's movement speed and direction are dynamically adjusted according to the position, speed and direction of the obstacle to perform local path planning and obstacle avoidance. The DWA algorithm quickly searches for feasible movement trajectories within the dynamic window of the robot's current position and selects the optimal trajectory to ensure that the robot moves safely and efficiently in complex dynamic environments. Step 4: Robot motion control. Based on the real-time adjusted path planning results, the computer outputs corresponding control commands to drive the robot's chassis, servo motors, and other actuators, controlling the robot to move along the optimized path and achieve autonomous exploration.

Images