Method and system for intelligent amphibious vehicle driving state safety perception based on multi-sensor fusion

Abstract

This invention discloses a method for safe perception of the driving status of an intelligent amphibious vehicle based on multi-sensor fusion, comprising: data acquisition, in which a central controller simultaneously acquires data from a depth sensor, gyroscope, and current sensor; data preprocessing, in which outliers and high-frequency noise are processed and standardized; multi-sensor data fusion and safety decision-making, in which the three types of data are correlated and analyzed to trigger buoyancy mode, water attitude control, and safety warning; and control execution and feedback, forming a closed-loop control. This invention, based on multi-sensor fusion, achieves accurate perception and collaborative data analysis of wading depth, vehicle attitude, and water flow environment through multi-sensor fusion technology. It enables intelligent triggering of buoyancy mode, dynamic stabilization of water attitude, and safety warning, thereby improving the driving safety of amphibious vehicles. This invention also discloses a system for safe perception of the driving status of an intelligent amphibious vehicle based on multi-sensor fusion.

Description

Technical Field

[0001] This invention belongs to the field of amphibious vehicle technology. Specifically, this invention relates to a method and system for safe perception of the driving status of an intelligent amphibious vehicle based on multi-sensor fusion. Background Technology

[0002] Amphibious vehicles, as special vehicles capable of operating in both land and water environments, face far greater complexity than ordinary land vehicles, demanding extremely high levels of safety awareness regarding their operational status. However, current technologies have several shortcomings in ensuring safety awareness during the operation of amphibious vehicles, as detailed below: 2.1 Insufficient accuracy in wading depth sensing and mode switching Current amphibious vehicles rely heavily on single-sensor water depth detection, and the sensor's installation location and sensing method are limited. For example, some vehicles mount water depth sensors on the bottom of the vehicle, which are easily damaged or have data distorted by collisions with mud, sand, or obstacles at the bottom. Some vehicles use a single type of water depth sensor (such as an infrared sensor), and the sensing accuracy drops significantly when there are waves, reflections, or floating objects (such as aquatic plants and plastic waste) on the water surface, with measurement errors reaching ±30cm or more. This directly leads to insufficient accuracy in determining whether the vehicle has entered buoyancy mode—if it misjudges (such as accidentally triggering buoyancy mode in shallow water), it will cause power system switching disorder and energy waste; if it misses (such as not triggering buoyancy mode in deep water), it may cause safety accidents due to excessive vehicle load or water entering the engine.

[0003] 2.2 Lack of vehicle attitude perception and stability control When amphibious vehicles navigate on water, their posture (such as tilt angle and sway amplitude) is easily altered by factors such as water flow, waves, and changes in their center of gravity. An imbalance in posture can lead to rollover. Current technology dictates that some amphibious vehicles lack dedicated attitude sensing equipment, relying solely on the driver's visual judgment of the vehicle's state, which is highly subjective and results in delayed responses. While some vehicles are equipped with attitude sensors, these are often single-axis or dual-axis sensors, unable to comprehensively acquire tilt angle values ​​and angular acceleration across the X, Y, and Z axes, making it difficult to accurately reflect real-time attitude changes. For example, when a vehicle tilts laterally under the impact of water flow, monitoring only the pitch angle while ignoring the roll angle fails to detect the risk of rollover in time, and prevents the triggering of appropriate attitude adjustment controls.

[0004] 2.3 Poor perception of water flow environment and navigation adaptability Water flow speed and direction are key environmental factors affecting the safety of amphibious vehicles in water navigation. Excessive flow speed can cause vehicles to be swayed off course by the current; an angle between the current and the navigation direction can lead to vehicle tilting or course deviation. Currently, most amphibious vehicles lack flow sensors, making them unable to perceive water flow conditions in real time. They rely solely on driver experience to adjust navigation direction and power output, resulting in extremely poor adaptability. For example, in waters with flow speeds exceeding 2 m / s, if the driver fails to detect changes in the current in time, the vehicle may be swept into shallows or rocky areas, or suffer from insufficient power due to excessive energy expenditure while fighting the current.

[0005] 2.4 Sensor data is isolated and lacks fusion applications. Even though some high-end amphibious vehicles are equipped with multiple sensors, these sensors often operate independently, outputting data in isolation without forming a collaborative perception mechanism. For example, water depth sensor data is only used to determine wading depth, gyroscope data is only used to display the vehicle's tilt angle, and flow velocity sensor data is not linked to vehicle control, resulting in wasted sensor resources and a one-sided perception of safety. This isolated mode cannot comprehensively assess the vehicle's overall safety status based on multi-dimensional information. For instance, if the water depth reaches the buoyancy mode trigger threshold, but the water flow velocity is too high, triggering buoyancy mode solely based on water depth data could cause the vehicle to lose control due to the impact of the water flow.

[0006] This paper provides a method for safety perception of the driving status of amphibious vehicles, specifically how to achieve accurate perception and collaborative data analysis of wading depth, vehicle attitude, and water flow environment through multi-sensor fusion technology. This solves the problems of low perception accuracy, insufficient attitude judgment, lack of water flow perception, and lack of data fusion, and ultimately realizes intelligent triggering of floating mode, dynamic stability of water attitude, and safety warning, thereby improving the driving safety of amphibious vehicles. Summary of the Invention

[0007] This invention aims to at least solve one of the technical problems existing in the prior art. To this end, this invention provides a method for safe perception of the driving status of intelligent amphibious vehicles based on multi-sensor fusion. The purpose is to achieve accurate perception and collaborative data analysis of wading depth, vehicle attitude, and water flow environment through multi-sensor fusion technology, thereby realizing intelligent triggering of floating mode, dynamic stabilization of water attitude, and safety early warning, and improving the driving safety of amphibious vehicles.

[0008] To achieve the above objectives, the technical solution adopted by this invention is: a method for safe perception of the driving status of an intelligent amphibious vehicle based on multi-sensor fusion, comprising the following steps: S1. Data Acquisition: The central controller synchronously acquires data from the depth sensor, gyroscope, and current sensor according to a set cycle. The depth sensor is used to obtain the vertical distance h from the sensor to the water surface, and calculates the actual wading depth by combining it with the vehicle's fixed height H. The depth sensors on the left and right sides collect redundant data. The gyroscope is used to acquire the tilt angle and angular acceleration of the XYZ axes. The current sensor is used to acquire the water flow velocity and the angle between the water flow direction and the vehicle's heading. S2. Data preprocessing: Outlier removal and smoothing filtering are performed on the collected data; S3. Multi-sensor data fusion and safety decision-making: Correlate and analyze the pre-processed data from the depth sensor, gyroscope, and current sensor to form a comprehensive judgment and trigger the corresponding control strategy. The control strategy includes at least a floating mode triggering strategy, a surface attitude control strategy, and a safety warning strategy. S4. Control Execution and Feedback: The central controller converts decisions into instructions and sends them to the vehicle's execution components. It also obtains the execution status through feedback sensors to form a closed-loop control.

[0009] In the data acquisition step, the set period is 100ms / time.

[0010] In the data preprocessing step, invalid and abnormal data are removed using the 3σ principle; moving average filtering is applied to the gyroscope and current sensor data, and the average of five consecutive samples is taken to reduce high-frequency noise; the data is standardized in a format, with water depth in meters and angles in degrees.

[0011] In the multi-sensor data fusion and safety decision-making steps, the floating mode triggering strategy is to trigger the floating mode when the following conditions are met simultaneously: water depth ≥ 1.2m for 1s, triaxial tilt angle ≤ 10° and angular acceleration ≤ 50° / s², flow velocity ≤ 2.5m / s and water flow angle ≤ 30°.

[0012] In the multi-sensor data fusion and safety decision-making steps, the water attitude control strategy is as follows: if the roll angle is greater than 8°, adjust the power of the two thrusters according to the direction of the water flow; if the pitch angle is greater than 6°, adjust the front and rear power distribution ratio according to the direction of the current.

[0013] In the multi-sensor data fusion and safety decision-making steps, the safety early warning strategy is as follows: when the water depth drops by more than 0.5m, the roll angle is greater than 15°, or the current velocity is greater than 4m / s, a graded early warning is triggered and the vehicle automatically decelerates and corrects its course.

[0014] The water depth sensor is an ultrasonic water depth sensor with a measurement range of 0-5m, an accuracy of ±1cm, a response time of ≤0.5s, and an IP68 waterproof rating. It connects to the central controller via an I2C interface. The gyroscope integrates a three-axis accelerometer and a three-axis gyroscope, with a measurement range of ±180° tilt angle, ±2000° / s angular acceleration, an accuracy of ±0.5°, and a sampling frequency of 100Hz. It connects to the central controller via an SPI interface. The current velocity sensor is an electromagnetic current velocity sensor with a measurement range of 0.05-10m / s, an accuracy of ±2%, and a response time of ≤1s. It connects to the central controller via an RS485 interface.

[0015] The central controller uses an Infineon TC275 microprocessor to simultaneously acquire data from three types of sensors, perform preprocessing, fusion, and decision control, and is connected to the central domain controller (HCU), the three types of sensors, and the vehicle actuators.

[0016] This invention also provides a safety perception system for the driving status of an intelligent amphibious vehicle based on multi-sensor fusion, comprising: The sensor module, including a water depth sensor, a gyroscope, and a current sensor, is used to collect water depth, vehicle attitude, and water flow environment data related to the amphibious vehicle's movement. The central controller is used to synchronously collect data from the sensor modules at a set period, preprocess the collected data, perform multi-sensor data fusion and safety decision-making, and convert the decisions into instructions to be sent to the vehicle execution components. At the same time, it receives the execution status feedback from the feedback sensors to form a closed-loop control. The vehicle execution component is used to receive instructions from the central controller and perform corresponding operations. The vehicle execution component includes multiple injection pumps and propellers.

[0017] The injection pump is located at the front of the vehicle, and the propeller is located at the rear of the vehicle.

[0018] This invention relates to a method for safe perception of the driving status of intelligent amphibious vehicles based on multi-sensor fusion. Through multi-sensor fusion technology, it achieves accurate perception and collaborative data analysis of wading depth, vehicle attitude, and water flow environment, solving the problems of low perception accuracy, insufficient attitude judgment, lack of water flow perception, and lack of data fusion. Ultimately, it can realize intelligent triggering of floating mode, dynamic stability of water attitude, and safety warning, thereby improving the driving safety of amphibious vehicles. Attached Figure Description

[0019] This manual includes the following figures, which illustrate the following: Figure 1 This is a diagram of the architecture of the intelligent amphibious vehicle driving status safety perception system of the present invention; Figure 2 This is a hardware-coordinated signal flow diagram. Detailed Implementation

[0020] The specific embodiments of the present invention will be further described in detail below with reference to the accompanying drawings, in order to help those skilled in the art to have a more complete, accurate and in-depth understanding of the concept and technical solutions of the present invention, and to facilitate its implementation.

[0021] In a first aspect, embodiments of the present invention provide a method for safety perception of the driving status of an intelligent amphibious vehicle based on multi-sensor fusion, comprising the following steps: S1. Data Acquisition: The central controller synchronously acquires data from the depth sensor, gyroscope, and current sensor according to a set cycle. The depth sensor is used to obtain the vertical distance h from the sensor to the water surface, and calculates the actual wading depth by combining it with the vehicle's fixed height H. The depth sensors on the left and right sides collect redundant data. The gyroscope is used to acquire the tilt angle and angular acceleration of the XYZ axes. The current sensor is used to acquire the water flow velocity and the angle between the water flow direction and the vehicle's heading. S2. Data preprocessing: Outlier removal and smoothing filtering are performed on the collected data; S3. Multi-sensor data fusion and safety decision-making: Correlate and analyze the pre-processed data from the depth sensor, gyroscope, and current sensor to form a comprehensive judgment and trigger the corresponding control strategy. The control strategy includes at least a floating mode triggering strategy, a surface attitude control strategy, and a safety warning strategy. S4. Control Execution and Feedback: The central controller converts decisions into instructions and sends them to the vehicle's execution components. It also obtains the execution status through feedback sensors to form a closed-loop control.

[0022] Specifically, the purpose of this invention is to achieve accurate perception and collaborative data analysis of wading depth, vehicle attitude, and water flow environment through multi-sensor fusion technology, thereby solving the problems of low perception accuracy, insufficient attitude judgment, lack of water flow perception, and lack of data fusion. Ultimately, it aims to achieve intelligent triggering of floating mode, dynamic stability of water attitude, and safety warning, thereby improving the driving safety of amphibious vehicles.

[0023] In this embodiment of the invention, in step S1 above, the sampling period for collecting data from the depth sensor, gyroscope, and flow velocity sensor is set to 100ms / time.

[0024] In this embodiment of the invention, the water depth sensor is an ultrasonic water depth sensor with a measurement range of 0-5m, an accuracy of ±1cm, a response time of ≤0.5s, and a waterproof rating of IP68 (stable operation from -20℃ to 80℃). It is connected to the central controller via an I2C interface. The gyroscope integrates a three-axis accelerometer and a three-axis gyroscope, with a measurement range of tilt angle ±180°, angular acceleration ±2000° / s, an accuracy of ±0.5°, and a sampling frequency of 100Hz. It is connected to the central controller via an SPI interface. The current velocity sensor is an electromagnetic current velocity sensor with a measurement range of 0.05-10m / s, an accuracy of ±2%, and a response time of ≤1s. It is connected to the central controller via an RS485 interface.

[0025] In this embodiment of the invention, the water depth sensor is fixed to the bottom of the left and right side rearview mirrors of the vehicle by a waterproof bracket made of 304 stainless steel. The waterproof bracket has a load-bearing capacity of ≥5kg, the water depth sensor is 5cm away from the bottom of the side rearview mirror, and the probe is vertically downward to ensure that it is perpendicular to the water surface.

[0026] In this embodiment of the invention, the gyroscope is fixed inside the vehicle's center of gravity control box. The gyroscope is placed horizontally and parallel to the vehicle's center axis. The control box is filled with shock-absorbing sponge to reduce vibration interference.

[0027] In this embodiment of the invention, the flow rate sensor is fixed to the bottom of the vehicle front by a sealing flange. The distance between the flow rate sensor and the bottom of the vehicle front is 30cm. The probe is submerged in water for 10cm and is always in contact with the water flow while floating. The probe surface is covered with a polytetrafluoroethylene anti-adhesion coating.

[0028] In this embodiment of the invention, the central controller uses an Infineon TC275 microprocessor to synchronously collect data from three types of sensors, perform preprocessing, fusion, and decision control, and is connected to the central domain controller (HCU), the three types of sensors, and the vehicle actuators, respectively.

[0029] In this embodiment of the invention, a vehicle actuator is used to receive instructions from a central controller and perform corresponding operations. The vehicle actuator includes a tire motor, a steering rudder, multiple injection pumps, and a propeller. Two injection pumps are located at the front of the vehicle. Two propellers are located at the rear of the vehicle, and each propeller includes a propeller.

[0030] In step S1 above, the central controller synchronously collects data from three types of sensors at a uniform cycle of 100ms / time: Water depth sensor: Measure the vertical distance h from the sensor to the water surface, and combine this with the fixed height H (e.g., 1.8m) from the exterior rearview mirror to the bottom of the vehicle to calculate the actual wading depth h. 深 = Hh, the left and right sensors collect data separately, forming redundant data; Gyroscope: Collects XYZ three-axis tilt angles (pitch angle: forward and backward tilt; roll angle: left and right tilt; yaw angle: heading deviation) and angular acceleration (reflects the rate of attitude change); Flow velocity sensor: Collects water flow velocity (m / s) and the angle between the water flow direction and the vehicle's heading. The angle between the water flow direction and the vehicle's heading is defined as 0° as downstream and 90° as lateral flow.

[0031] In this embodiment of the invention, in step S2 above, the collected data is denoised and standardized to ensure data reliability. When removing outliers, the 3σ principle (data deviating from the average by more than 3 times the standard deviation is considered outlier) is used to remove invalid outlier data, such as the "0 value" of a depth sensor due to wave mismeasurement.

[0032] In this embodiment of the invention, during the smoothing filtering in step S2, a moving average filter is used for the gyroscope and flow velocity sensor data, and the average value of five consecutive samples is taken to reduce high-frequency noise, such as flow velocity fluctuations caused by water flow fluctuations.

[0033] In this embodiment of the invention, in step S2 above, the data is standardized in a uniform format, with water depth in meters and angle in degrees, to facilitate subsequent data fusion.

[0034] In this embodiment of the invention, in step S3 above, a comprehensive judgment is formed and a control strategy is triggered by analyzing three types of data: The floating mode trigger strategy is as follows: When the following conditions are met simultaneously: water depth ≥ 1.2m for 1s, triaxial tilt angle ≤ 10° and angular acceleration ≤ 50° / s², flow velocity ≤ 2.5m / s and water flow angle ≤ 30°, the floating mode is triggered to avoid misjudgment based on a single condition. The water attitude control strategy is as follows: if the roll angle is greater than 8°, adjust the power of the two propellers according to the direction of the water flow. For example, when the right side is tilted, increase the propeller speed of the propeller arranged on the left side; if the pitch angle is greater than 6°, adjust the front and rear power distribution ratio according to the direction of the current.

[0035] In the multi-sensor data fusion and safety decision-making process, the safety warning strategy is as follows: when the water depth drops by more than 0.5m, the roll angle is greater than 15°, or the current velocity is greater than 4m / s, a graded warning is triggered, and the vehicle automatically decelerates and corrects its course. A water depth drop of more than 0.5m indicates a possible bottoming out. A roll angle greater than 15° indicates a risk of capsizing. A current velocity greater than 4m / s indicates extreme danger, and the vehicle needs to decelerate.

[0036] Secondly, embodiments of the present invention also provide an intelligent amphibious vehicle driving status safety perception system based on multi-sensor fusion, comprising: The sensor module, including a water depth sensor, a gyroscope, and a current sensor, is used to collect water depth, vehicle attitude, and water flow environment data related to the amphibious vehicle's movement. The central controller is used to synchronously collect data from the sensor modules according to a set period, preprocess the collected data, fuse multi-sensor data and make safety decisions, and convert the decisions into instructions to send to the vehicle's execution components. At the same time, it receives execution status feedback from the feedback sensors to form a closed-loop control. The vehicle actuator is used to receive instructions from the central controller and perform corresponding operations. The vehicle actuator includes multiple injection pumps and propellers.

[0037] The intelligent amphibious vehicle driving status safety perception method and system based on multi-sensor fusion of the present invention has the following advantages: (1) Significantly improved sensing accuracy: water depth measurement error decreased from ±30cm to ±5cm, attitude tilt angle error decreased from ±2° to ±0.5°, and water flow velocity error decreased from ±10% to ±2%, providing a reliable basis for safety decision-making; (2) More reliable floating mode switching: Based on multiple conditions such as water depth, attitude and water flow, the switching accuracy has been improved from 75% to over 98%, avoiding safety risks caused by false triggering / missed triggering; (3) More stable navigation attitude: In waters with a current speed of 2-3 m / s, the side roll angle can be controlled within ±5°, the heading deviation is ≤±3°, and the stability is improved by more than 60%; (4) The vehicle has a stronger environmental adaptability: it can drive safely in complex waters (shallows, rapids, etc.) with a flow rate of 0.05-3m / s, and its applicable range has been greatly expanded; (5) More comprehensive safety: Construct a full-process mechanism of prevention (early detection of danger) - control (dynamic adjustment of posture) - early warning (gradual prompts to drivers) to reduce the risk of vehicle driving in all aspects.

[0038] The present invention has been described above by way of example with reference to the accompanying drawings. Obviously, the specific implementation of the present invention is not limited to the above-described manner. Any non-substantial improvements made using the inventive concept and technical solution; or the direct application of the inventive concept and technical solution to other situations without modification, are all within the protection scope of the present invention.

Claims

1. A method for safety perception of the driving status of intelligent amphibious vehicles based on multi-sensor fusion, characterized in that, Includes the following steps: S1. Data Acquisition: The central controller synchronously acquires data from the depth sensor, gyroscope, and current sensor according to a set cycle. The depth sensor is used to obtain the vertical distance h from the sensor to the water surface, and calculates the actual wading depth by combining it with the vehicle's fixed height H. The depth sensors on the left and right sides collect redundant data. The gyroscope is used to acquire the tilt angle and angular acceleration of the XYZ axes. The current sensor is used to acquire the water flow velocity and the angle between the water flow direction and the vehicle's heading. S2. Data preprocessing: Outlier removal and smoothing filtering are performed on the collected data; S3. Multi-sensor data fusion and safety decision-making: Correlate and analyze the pre-processed data from the depth sensor, gyroscope, and current sensor to form a comprehensive judgment and trigger the corresponding control strategy. The control strategy includes at least a floating mode triggering strategy, a surface attitude control strategy, and a safety warning strategy. S4. Control Execution and Feedback: The central controller converts decisions into instructions and sends them to the vehicle's execution components. It also obtains the execution status through feedback sensors to form a closed-loop control.

2. The method for safety perception of the driving status of an intelligent amphibious vehicle based on multi-sensor fusion according to claim 1, characterized in that, In the data acquisition step, the set period is 100ms / time.

3. The method for safety perception of the driving status of an intelligent amphibious vehicle based on multi-sensor fusion according to claim 1, characterized in that, In the data preprocessing step, invalid and abnormal data are removed using the 3σ principle; moving average filtering is applied to the gyroscope and current sensor data, and the average of five consecutive samples is taken to reduce high-frequency noise; the data is standardized in a format, with water depth in meters and angles in degrees.

4. The method for safety perception of the driving status of an intelligent amphibious vehicle based on multi-sensor fusion according to any one of claims 1 to 3, characterized in that, In the multi-sensor data fusion and safety decision-making steps, the floating mode triggering strategy is to trigger the floating mode when the following conditions are met simultaneously: water depth ≥ 1.2m for 1s, triaxial tilt angle ≤ 10° and angular acceleration ≤ 50° / s², flow velocity ≤ 2.5m / s and water flow angle ≤ 30°.

5. The method for safety perception of the driving status of an intelligent amphibious vehicle based on multi-sensor fusion according to any one of claims 1 to 3, characterized in that, In the multi-sensor data fusion and safety decision-making steps, the water attitude control strategy is as follows: if the roll angle is greater than 8°, adjust the power of the two thrusters according to the direction of the water flow; if the pitch angle is greater than 6°, adjust the front and rear power distribution ratio according to the direction of the current.

6. The method for safety perception of the driving status of an intelligent amphibious vehicle based on multi-sensor fusion according to any one of claims 1 to 3, characterized in that, In the multi-sensor data fusion and safety decision-making steps, the safety early warning strategy is as follows: when the water depth drops by more than 0.5m, the roll angle is greater than 15°, or the current velocity is greater than 4m / s, a graded early warning is triggered and the vehicle automatically decelerates and corrects its course.

7. The method for safety perception of the driving status of an intelligent amphibious vehicle based on multi-sensor fusion according to any one of claims 1 to 3, characterized in that, The water depth sensor is an ultrasonic water depth sensor with a measurement range of 0-5m, an accuracy of ±1cm, a response time of ≤0.5s, and an IP68 waterproof rating. It connects to the central controller via an I2C interface. The gyroscope integrates a three-axis accelerometer and a three-axis gyroscope, with a measurement range of ±180° tilt angle, ±2000° / s angular acceleration, an accuracy of ±0.5°, and a sampling frequency of 100Hz. It connects to the central controller via an SPI interface. The current velocity sensor is an electromagnetic current velocity sensor with a measurement range of 0.05-10m / s, an accuracy of ±2%, and a response time of ≤1s. It connects to the central controller via an RS485 interface.

8. The method for safety perception of the driving status of an intelligent amphibious vehicle based on multi-sensor fusion according to any one of claims 1 to 3, characterized in that, The central controller uses an Infineon TC275 microprocessor to simultaneously acquire data from three types of sensors, perform preprocessing, fusion, and decision control, and is connected to the central domain controller (HCU), the three types of sensors, and the vehicle actuators.

9. A safety perception system for the driving status of an intelligent amphibious vehicle based on multi-sensor fusion, characterized in that, include: The sensor module, including a water depth sensor, a gyroscope, and a current sensor, is used to collect water depth, vehicle attitude, and water flow environment data related to the amphibious vehicle's movement. The central controller is used to synchronously collect data from the sensor modules at a set period, preprocess the collected data, perform multi-sensor data fusion and safety decision-making, and convert the decisions into instructions to be sent to the vehicle execution components. At the same time, it receives the execution status feedback from the feedback sensors to form a closed-loop control. The vehicle execution component is used to receive instructions from the central controller and perform corresponding operations. The vehicle execution component includes multiple injection pumps and propellers.

10. The intelligent amphibious vehicle driving status safety perception system based on multi-sensor fusion according to claim 9, characterized in that, The injection pump is located at the front of the vehicle, and the propeller is located at the rear of the vehicle.

Images