Auxiliary adjustment method for wall-attached walking posture of underwater robot

Abstract

The invention discloses an underwater robot wall-attached walking attitude auxiliary adjustment method, and relates to the technical field of underwater robots, and the method comprises the steps: obtaining a plurality of groups of altimeter data and inertial navigation data, and processing the altimeter data and inertial navigation data to obtain optimized altimeter data; comparing differences among the plurality of groups of optimized altimeter data, and when the differences are smaller than a threshold value, executing wall attachment preparation operation; calculating a roll angle and a pitch angle of the robot by using data differences of the altimeters and geometric parameters of the robot; the attitude of the underwater robot is accurately controlled through fusion processing of the altimeter and inertial navigation data, and stable wall-attached walking and operation execution are achieved.

Description

Technical Field

[0001] The invention belongs to the technical field of underwater robots, and in particular relates to an auxiliary adjustment method for an underwater robot's wall-sticking walking posture. Background Art

[0002] In the prior art, underwater robots face a series of challenges, especially when performing non-ground maintenance tasks such as facades, inclined surfaces and top surfaces. Among them, wall walking and posture control are the two core problems. Wall walking requires the underwater robot to be able to stably approach the maintenance surface in order to carry out detailed inspection and maintenance work. However, due to the instability of the water flow, the fluctuation of the robot's own power and the interference of environmental factors, the robot often deviates, slides or rotates during the wall walking process. This not only affects the accuracy and efficiency of the maintenance, but may also cause damage to the robot itself and the maintenance surface.

[0003] To solve this problem, operators usually need to spend a lot of time and energy to make manual adjustments. They need to constantly observe the robot's posture and position and manipulate the control device to adjust the robot's propulsion, steering, and posture. However, this manual adjustment method is not only inefficient, but also requires extremely high operator skills. In complex underwater environments, even experienced operators find it difficult to ensure that adjustments can be completed quickly and accurately every time.

[0004] Posture control is another key challenge. In vertical, inclined and top surface maintenance tasks, underwater robots need to maintain a specific posture to adapt to the shape and angle of the maintenance surface. However, due to various factors, it is often difficult for robots to maintain a stable posture in actual operations. This may cause the maintenance tool to be unable to accurately align with the maintenance surface, thus affecting the quality and efficiency of the maintenance. Summary of the invention

[0005] The technical problem to be solved by the present invention is to provide an auxiliary adjustment method for the underwater robot's wall-walking posture. By acquiring and processing multiple sets of altimeter data and inertial navigation data, the roll angle and pitch angle of the underwater robot can be more accurately calculated, thereby achieving more accurate wall-walking. At the same time, the PID controller is used to calculate the control output according to the posture error, so that the robot's posture can be adjusted in time to maintain the stability of wall-walking.

[0006] In order to solve the above technical problems, the technical solution adopted by the present invention is: An auxiliary adjustment method for an underwater robot's wall-climbing walking posture, comprising the following steps: Acquire multiple sets of altimeter data and inertial navigation data, and process the altimeter data and inertial navigation data to obtain optimized altimeter data; Compare the differences between the optimized multiple groups of altimeter data, and when the difference is less than a threshold, perform a wall-attaching preparation operation; Using the difference in altimeter data and the robot's geometric parameters, the robot's roll and pitch angles are calculated; Fuse the angles calculated by the altimeter data and the inertial navigation data to obtain the fused attitude angle; Determine the attitude error according to the difference between the desired attitude angle and the fused attitude angle, and use the PID controller to calculate the control output to control the attitude of the underwater robot; The fused data is used to maintain the wall-adhering posture using a PID controller to achieve wall-adhering walking and perform tasks.

[0007] Further, multiple sets of altimeter data and inertial navigation data are obtained, and the altimeter data and inertial navigation data are processed to obtain optimized altimeter data, including: Acquire multiple sets of altitude data through the altimeter sensor, and acquire acceleration and angular velocity data through the inertial navigation system; Filtering the altimeter data to obtain smoothed altimeter data, and calibrating the inertial navigation data to obtain calibrated acceleration and angular velocity data; Establish the state equation based on the inertial navigation data, and use the altimeter data as the observation value to establish the observation equation; According to the altimeter data and inertial navigation data, the initial state estimate is set, and the prediction error covariance matrix and the process and observation noise covariance matrix are initialized; Use the state estimate and state equation to predict the current state and update the prediction error covariance matrix. When new altimeter data arrives, calculate the Kalman gain. Update the state estimate using the Kalman gain and new observations to fuse the prediction with the observation information and update the prediction error covariance matrix; Continuously iterate and update the status every time new altimeter data arrives; When the preset number of iterations is reached, the filter outputs the optimized height data.

[0008] Furthermore, the calculation formula of the state equation is: ; in, represents the state vector at time k; represents the state transfer matrix; Indicates at the last time point The state vector of represents the control input matrix; Indicated in Control input at all times; represents process noise.

[0009] Furthermore, the calculation formula of the observation equation is: ; in, Indicates at time Observed value of represents the observation matrix; represents the observation noise.

[0010] Furthermore, the calculation formula of Kalman gain is: ; in, Indicates at time The Kalman gain of represents the prediction error covariance matrix; represents the transpose of the observation matrix; represents the observation noise covariance matrix.

[0011] Furthermore, the calculation formula of the fused posture angle is: ; in, Indicates the posture angle after fusion; Represents the attitude angle calculated based on the difference in altimeter data; Indicates the attitude angle calculated based on the inertial navigation data; Represents the dynamically adjusted fusion weight.

[0012] Further, the attitude error is determined according to the difference between the desired attitude angle and the fused attitude angle, and a PID controller is used to calculate the control output to control the attitude of the underwater robot, including: Set the parameters of the PID controller, including the proportional coefficient, integral coefficient and differential coefficient; Initialize the integral term and the last error of the PID controller to 0, and obtain the preset desired attitude angle and the fused current attitude angle to obtain the attitude error; According to the attitude error, the proportional, integral and differential control components are calculated by the PID controller to obtain the total output of the PID controller; The total output of the PID controller is converted into control instructions to control the posture of the underwater robot.

[0013] An auxiliary adjustment system for an underwater robot walking against a wall, comprising: The acquisition module is used to acquire multiple sets of altimeter data and inertial navigation data, and process the altimeter data and inertial navigation data to obtain optimized altimeter data; compare the differences between the optimized multiple sets of altimeter data, and when the difference is less than a threshold, perform a wall-attaching preparation operation; and calculate the roll angle and pitch angle of the robot using the altimeter data difference and the robot's geometric parameters; The processing module is used to fuse the angles calculated by the altimeter data and the inertial navigation data to obtain the fused attitude angle; determine the attitude error according to the difference between the expected attitude angle and the fused attitude angle, and use the PID controller to calculate the control output to control the attitude of the underwater robot; maintain the wall-adhering attitude according to the fused data and use the PID controller to achieve wall-adhering walking and perform operations.

[0014] A computing device comprising: one or more processors; The storage device is used to store one or more programs. When the one or more programs are executed by the one or more processors, the one or more processors implement the method described.

[0015] In a fourth aspect, a computer-readable storage medium stores a program, and when the program is executed by a processor, the method described is implemented.

[0016] The present invention can achieve the following beneficial effects: 1. By acquiring and processing multiple sets of altimeter data and inertial navigation data, the roll angle and pitch angle of the underwater robot can be calculated more accurately, thereby achieving more accurate wall-clinging walking. At the same time, the PID controller is used to calculate the control output according to the attitude error, so that the robot's attitude can be adjusted in time to maintain the stability of wall-clinging walking.

[0017] 2. Process the altimeter data and inertial navigation data to obtain optimized altimeter data, thereby reducing data errors and noise interference and improving data reliability and accuracy. This helps the robot achieve more reliable attitude control in complex and changeable underwater environments.

[0018] 3. By comparing the differences between the optimized sets of altimeter data, the wall preparation operation is performed when the difference is less than the threshold, which can reduce invalid wall attempts, save time and improve work efficiency. At the same time, the use of PID controller to quickly adjust the attitude error can also shorten the adjustment time and improve work efficiency. Through automated data processing and attitude adjustment, the requirements for operator skills are reduced. The operator does not need to spend a lot of time and energy on manual adjustments, which reduces the workload and improves work comfort. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] The present invention will be further described below in conjunction with the accompanying drawings and embodiments: Figure 1 A schematic flow chart of a method for assisting adjustment of the wall-walking posture of an underwater robot according to the present invention; Figure 2 The present invention is a schematic diagram of an auxiliary adjustment system for an underwater robot's wall-walking posture. DETAILED DESCRIPTION

[0020] The preferred solution is Figure 1 to Figure 2 As shown, a method for assisting adjustment of the posture of an underwater robot walking against a wall is provided. Step 1, acquiring multiple sets of altimeter data and inertial navigation data, and processing the altimeter data and the inertial navigation data to obtain optimized altimeter data; Step 2, comparing the differences between the optimized multiple groups of altimeter data, and when the difference is less than a threshold, performing a wall-attaching preparation operation; Step 3, using the difference in altimeter data and the robot's geometric parameters, calculate the robot's roll angle and pitch angle; Step 4, fusing the angles calculated by the altimeter data and the inertial navigation data to obtain the fused attitude angle; Step 5, determining the attitude error according to the difference between the desired attitude angle and the fused attitude angle, and using a PID controller to calculate the control output to control the attitude of the underwater robot; Step 6: Based on the fused data, a PID controller is used to maintain the wall-sticking posture to achieve wall-sticking walking and perform the task.

[0021] In an embodiment of the present invention, by acquiring multiple sets of altimeter data and inertial navigation data, and processing and optimizing these data, the accuracy and reliability of the data can be improved. The optimized altimeter data provides a more reliable basis for subsequent posture calculation and wall-sticking walking, and helps to reduce the interference of errors and noise. Compare the differences between the optimized multiple sets of altimeter data, and perform wall-sticking preparatory operations when the differences are less than the set threshold. This step can ensure that the robot has reached a relatively stable posture before sticking to the wall. This prejudgment mechanism helps to improve the success rate and efficiency of wall-sticking walking, and reduce invalid attempts and time waste. Using the altimeter data differences and the robot's geometric parameters to calculate the robot's roll angle and pitch angle is relatively simple and practical.

[0022] The angles calculated by the altimeter data and the inertial navigation data are fused to obtain the fused attitude angle. This step comprehensively utilizes the information of the two data sources and improves the accuracy and robustness of the attitude angle. This data fusion method helps the robot achieve more stable attitude control in complex and changeable underwater environments. The attitude error is determined by comparing the difference between the expected attitude angle and the fused attitude angle, and the control output is calculated using the PID controller. This step realizes the precise control of the underwater robot's attitude. The PID controller can quickly adjust the robot's attitude according to the attitude error so that it can reach the expected attitude as soon as possible, thereby improving the accuracy and efficiency of the operation. According to the fused data and using the PID controller to maintain the wall-sticking attitude to achieve wall-sticking walking and perform operations, this step ensures the stability and accuracy of the robot during the wall-sticking walking process. By continuously monitoring and adjusting the attitude, the robot can closely fit the maintenance surface for operation, improving the quality and efficiency of maintenance. At the same time, this automated control method also reduces the burden and skill requirements of the operator.

[0023] In a preferred embodiment of the present invention, the above step 1, obtaining multiple sets of altimeter data and inertial navigation data, and processing the altimeter data and inertial navigation data to obtain optimized altimeter data, includes: Step 11, obtaining multiple sets of altitude data through an altimeter sensor, and obtaining acceleration and angular velocity data through an inertial navigation system; Step 12, filtering the altimeter data to obtain smoothed altimeter data, and calibrating the inertial navigation data to obtain calibrated acceleration and angular velocity data; Step 13, establishing a state equation based on the inertial navigation data, and using the altimeter data as observation values ​​to establish an observation equation; Step 14, according to the altimeter data and the inertial navigation data, set the initial state estimate, initialize the prediction error covariance matrix and define the process and observation noise covariance matrix; Step 15, predict the current state using the state estimate and the state equation, and update the prediction error covariance matrix. When new altimeter data arrives, calculate the Kalman gain; Step 16, using the Kalman gain and the new observation data to update the state estimate to fuse the prediction with the observation information, and update the prediction error covariance matrix; Step 17, continue iterating and updating, updating the status each time new altimeter data arrives; Step 18: When the preset number of iterations is reached, the filter outputs the optimized height data.

[0024] In an embodiment of the present invention, an altimeter sensor is used to continuously capture multiple sets of altitude data, which reflect the real-time distance between the underwater robot and the wall. At the same time, the inertial navigation system continuously provides the robot's acceleration and angular velocity data to reveal the robot's dynamic behavior. The altimeter data is processed by a filtering algorithm (such as a low-pass filter) to remove noise and instantaneous interference and ensure the smoothness of the data. The inertial navigation data undergoes a calibration step to correct the deviation of the sensor itself and obtain more accurate acceleration and angular velocity readings. Based on the inertial navigation data, state equations are constructed, which describe the evolution of state variables such as the robot's position and velocity. The processed altimeter data is used as the observation value, and the observation equation is established based on it to associate the observation data with the robot state. The initial state estimate is set, the prediction error covariance matrix is ​​initialized, and the uncertainty of the initial state estimate is evaluated. The covariance matrix of the process and observation noise is defined to reflect the error characteristics of the sensor and the model.

[0025] The state equation and the current state estimate are used to predict the future state and update the covariance matrix of the prediction error. Whenever new altitude data arrives, the Kalman gain is calculated in real time based on the prediction error covariance matrix, the observation noise covariance matrix and the observation matrix. The state estimate is updated using the newly calculated Kalman gain and the latest observation data to achieve the optimal fusion of prediction and observation information. The prediction error covariance matrix is ​​updated to reflect changes in the accuracy of the state estimate. As new data continues to arrive, the above prediction and update steps are repeated to achieve dynamic optimization of the state. When the number of iterations reaches the preset value or meets specific conditions, the filter will output the optimized altitude data.

[0026] Through filtering and calibration processing, the noise and deviation in the original data are effectively eliminated, and the accuracy of the height data and inertial navigation data is significantly improved. The recursive characteristics of the Kalman filter make the data gradually stabilize during the processing process, reducing the impact of sudden interference on the system. The information from different sensors is integrated, the advantages of each sensor are fully utilized, and the reliability and robustness of the data are improved. With the continuous input of new data, the system can adjust the state estimation in real time to adapt to the changes in the robot's motion state and maintain the real-time and effectiveness of the data. The optimized height data provides a more precise control basis for the underwater robot's wall-clinging walking and posture adjustment, which helps to improve the robot's operating efficiency and safety.

[0027] In a preferred embodiment of the present invention, the calculation formula of the state equation is: ; in, represents the state vector at time k; represents the state transfer matrix; Indicates at the last time point The state vector of represents the control input matrix; Indicated in Control input at all times; represents process noise.

[0028] In an embodiment of the present invention, Indicates that the system is at time The complete state of a system is a vector that contains all the variables needed to describe the dynamics of the system. Over time, the state vector is updated based on the dynamic behavior of the system. is a matrix that describes how the system state transfers from one time point to the next in the absence of external input and noise interference. The elements of the matrix are determined based on physical models or system identification techniques, reflecting the dynamic characteristics of the system. Matrix and the state vector at the previous time point By multiplying them, we can predict that in the absence of external control input and noise, the system status. Represents the previous point in time The state of the system. It is used in the state equation to calculate the current state Based on the state transfer matrix By performing operations on the system, we can deduce the natural evolution of the system without external influences. The matrix describes how the control input affects the change of the system state. It is converted into the impact on the state vector, reflecting the response characteristics of the system to external control.

[0029] The dimensions and structure of the matrices match the dimensions of the state vector and control input vector to ensure correct mathematical operations. Indicated in Control input is any external signal or operation that can affect the behavior of the system, such as the acceleration command of the robot, the steering angle of the servo, etc. By multiplying the matrices, the control input is converted into a direct effect on the system state. Represents the random noise or disturbance introduced during the state transition process. Process noise is inevitable and may come from many factors, such as sensor error, environmental disturbance or model uncertainty.

[0030] The purpose of introducing process noise into the state equation is to simulate the dynamic behavior of the actual system more realistically.

[0031] Provides a specific point in time A complete description of the system status helps to fully understand the current status of the system. Knowing the current status is the basis for predicting the future behavior of the system and making control decisions. The matrix effectively captures the natural evolution of the system state over time and can predict the future state of the system without external input. The eigenvalues ​​and eigenvectors of the matrix can be used to evaluate the stability of the system. It represents the historical state of the system and provides background information for the calculation of the current state. Through the state vector at consecutive time points, the continuous changes of the system state can be tracked. The matrix quantifies the direct effect of control input on the change of system state, providing key information for the design of control systems. Understanding how control input affects system state helps to develop more effective control strategies. , the system can be actively controlled and manipulated to achieve a desired state or behavior. The presence of control inputs enables the system to adapt to different operating conditions and task requirements. The inevitable uncertainty in the actual system is taken into account, making the model closer to reality.

[0032] In a preferred embodiment of the present invention, the calculation formula of the observation equation is: ; in, Indicates at time Observed value of represents the observation matrix; represents the observation noise.

[0033] In an embodiment of the present invention, Represents the observation data directly obtained by sensors or other measuring devices at time k. The observation value is a partial or direct reflection of the system state. is a matrix that establishes the system state With observed value The observation matrix describes how to map from the system state space to the observation space, that is, it defines which state variables can be directly observed and how to extract these observation information from the state variables. In some cases, is an identity matrix or selection matrix representing a subset of directly observed state variables, Indicates that the system is at time The true state of the system is a vector containing all relevant states of the system. In the observation equation, is the object being observed, that is, we hope to observe the value to estimate or restore the state. represents the random noise or error introduced during the observation process. Observation noise may come from many factors such as sensor imperfections, environmental interference, quantization errors, etc. In the observation equation, is a random variable used to model the deviation between the observed value and the actual state.

[0034] It provides direct measurement data of certain aspects of the system, which is an important basis for verifying and updating the system state estimation. The observation value is usually obtained in real time at a specific point in time, so it can reflect the actual current state of the system, which helps to find problems and adjust the control strategy in time. The matrix establishes the mathematical relationship between the system state and the observation value, making it possible to map from the state space to the observation space. In some cases, the observation matrix can play a role in dimensionality reduction or selection of specific state variables, thereby simplifying the processing and analysis of observation data. It contains all the necessary information to describe the state of the system and is the basis of the observation equation.

[0035] An accurate state vector is a prerequisite for effective state estimation, and the observation equation provides a reverse mapping from the observation value to the state vector. Taking into account the inevitable uncertainty in the observation process makes the observation equation more consistent with the actual situation and improves the robustness of the model. By explicitly considering the observation noise, appropriate methods (such as filtering algorithms) can be used in the state estimation process to reduce the impact of noise and improve the estimation accuracy.

[0036] In a preferred embodiment of the present invention, the calculation formula of the Kalman gain is: ; in, Indicates at time The Kalman gain of represents the prediction error covariance matrix; represents the transpose of the observation matrix; represents the observation noise covariance matrix.

[0037] In the embodiment of the present invention, the Kalman gain is a weighting factor used to combine the forecast information with the observation information in the update step. It determines the importance of the new observation data in revising the state estimate. A larger gain value means that the observation data has a greater impact on the state update, while a smaller gain value means that the forecast data has a greater impact. The forecast error covariance matrix represents the uncertainty in the state prediction. After each prediction step, this matrix is ​​updated to reflect the uncertainty in the state prediction due to system dynamics and process noise. The size of will affect the calculation of Kalman gain, and the larger Values ​​of 1 will result in a larger Kalman gain, meaning that the observations will have more weight in the update step. H Describes the relationship between state variables and observations. In the calculation of Kalman gain, ( H The transpose of ) is used to transform the information in the observation space into the state space. H and The role of is to link the state estimate with the observation data so that the observation data can be used to update the state estimate. R Represents the statistical characteristics of observation noise, that is, the uncertainty in the observation data. It reflects the reliability or accuracy of the observation data. R A value of means that the observed data contains more noise or uncertainty, so it should be given a smaller weight when updating the state. In the calculation of the Kalman gain, R The prediction error covariance matrix is ​​obtained by Added together to balance the weights of prediction and observation information.

[0038] The Kalman gain can dynamically adjust the weights of the predicted and observed data in the state estimation to ensure that the optimal state estimation can be obtained at any time. The calculated Kalman gain can effectively combine the new observation data with the previous prediction information to improve the accuracy and reliability of the state estimation. Being able to quantify the uncertainty of state predictions helps to understand the credibility of current state estimates. Through this matrix, the risk of predictions can be evaluated so that these risk factors can be considered in subsequent decision-making and planning. and its transpose The connection between state variables and observations is established, so that the observation data can be directly used to update the state estimate. It can effectively convert the information in the observation space to the state space, which is convenient for subsequent data processing and state update. R , the noise in the observation data can be modeled and quantified, thereby improving the robustness of the state estimation. This matrix helps to reasonably allocate the weight of the observation data during the state update process, avoiding excessive impact of data with excessive noise on the state estimation.

[0039] In a preferred embodiment of the present invention, the above step 2, comparing the differences between the optimized multiple sets of altimeter data, and when the difference is less than a threshold, performing a wall-attaching preparation operation, may include: Collect raw data from multiple altimeters simultaneously. Preprocess the raw data, including noise removal, outlier detection and removal, data smoothing, etc., to optimize data quality. Ensure that the data from all altimeters are synchronized in time for accurate comparison. If the data acquisition frequency is different, the data needs to be aligned to the same time point through interpolation or other methods. Compare the optimized multiple groups of altimeter data point by point and calculate the difference between them. The difference can be obtained by calculating the absolute difference or relative difference between the two groups of data. Set a difference threshold, which is determined based on the actual application requirements and system tolerance. Compare the calculated difference with the set threshold. If the difference of all groups of data is less than the set threshold, the data between the altimeters is considered to be consistent. Once the difference is detected to be less than the threshold, the decision logic for the wall preparation operation is triggered. When the difference is less than the threshold, the system starts to perform the wall preparation operation.

[0040] The wall preparation operation may include adjusting the device posture, preparing the contact surface, activating related sensors or actuators, etc. These operations are designed to ensure that the device can smoothly and accurately contact the wall during the subsequent wall preparation process. During the wall preparation operation, continuously monitor the changes in the altimeter data. If the data is abnormal or the difference exceeds the threshold, immediately interrupt the wall preparation operation and make necessary adjustments, including recalibrating the altimeter, checking the device status, or modifying the operating parameters. Record key data points throughout the process, including altimeter readings, difference calculation results, and the execution of wall preparation operations.

[0041] In a preferred embodiment of the present invention, the above step 3, using the altimeter data difference and the robot's geometric parameters to calculate the robot's roll angle and pitch angle, may include: Collect data from multiple altimeters installed on the robot. At least three altimeters that are not in the same straight line are required to accurately calculate the roll and pitch angles. Filter and smooth the collected altimeter data to reduce the influence of noise and outliers. According to the data differences between different altimeters, calculate the corresponding height difference. For example, the height difference between altimeter 1 and altimeter 2 is calculated as: ; Similarly, the altitude difference between altimeter 1 and altimeter 3 can be calculated as: ; and the altitude difference between altimeter 2 and altimeter 3 is: .

[0042] Assume that the three altimeters on the robot are installed at the three vertices of a triangle, forming a fixed geometric relationship. Set up a coordinate system, where the origin is located at a vertex of the triangle. X Axis and YThe axes are parallel to the two sides of the triangle. The height difference of each vertices of the triangle is used, combined with the geometric parameters of the robot (such as the length of the side of the triangle, the relative position between the altimeters, etc.), to calculate the robot's roll angle and pitch angle through trigonometric function relationships.

[0043] For example, the specific implementation process is as follows: , and , and the horizontal distance between them is known. For the roll angle roll , calculated using the height difference between two adjacent altimeters and the horizontal distance between them: ;in, is the horizontal distance between altimeter 1 and altimeter 2. For the pitch angle pitch , using the data from the other two altimeters to calculate: ;in, is the horizontal distance between altimeter 1 and altimeter 3. According to the actual situation, the calculated angle is corrected to eliminate the installation error or measurement error. The calculated roll angle and pitch angle are output to the robot control system.

[0044] In a preferred embodiment of the present invention, in the above step 4, the angle calculated by the altimeter data and the inertial navigation data is fused to obtain a fused attitude angle; the calculation formula of the fused attitude angle is: ; in, Indicates the posture angle after fusion; Represents the attitude angle calculated based on the difference in altimeter data; Indicates the attitude angle calculated based on the inertial navigation data; Represents the dynamically adjusted fusion weight.

[0045] In an embodiment of the present invention, real-time data is obtained from an altimeter installed on a robot or device, which reflects the height change of the device relative to the ground. Data is obtained from inertial sensors such as gyroscopes and accelerometers, which provide angular velocity and acceleration information of the device. According to the difference in altimeter data, the attitude angle is calculated using geometric relationships and trigonometric functions. It involves comparing the readings of altimeters at different locations to estimate the roll angle and pitch angle. Inertial navigation attitude angle The change in attitude angle is estimated by integrating the gyroscope data (angular velocity), and may be combined with accelerometer data for initial alignment and correction of attitude.

[0046] It is a dynamically adjusted weight factor used to determine the relative importance of altimeter data and inertial navigation data in the fusion process. The value of can be adjusted according to different conditions, such as the reliability of the sensor, the degree of dynamic environment, signal quality, etc.

[0047] For example, in static or low-dynamic situations, altimeter data may be more reliable, so can be set higher; in highly dynamic environments, inertial navigation data may be more accurate, so Should be reduced accordingly.

[0048] Using the formula The attitude angles calculated from the two data sources are fused. This fusion process combines the static accuracy of the altimeter and the dynamic response capability of the inertial navigation system to obtain a more accurate and stable attitude estimate. is output to the control system, At the same time, the fusion algorithm can be adjusted according to the actual application situation to optimize Dynamic adjustment strategy.

[0049] In a preferred embodiment of the present invention, the above step 5, determining the posture error according to the difference between the desired posture angle and the fused posture angle, and using the PID controller to calculate the control output to control the posture of the underwater robot, may include: Step 51, setting the parameters of the PID controller, including the proportional coefficient, the integral coefficient and the differential coefficient; Step 52, initializing the integral term and the last error of the PID controller to 0, and obtaining the preset desired attitude angle and the fused current attitude angle to obtain the attitude error; Step 53, according to the posture error, the proportional, integral and differential control components are calculated by the PID controller to obtain the total output of the PID controller; Step 54, converting the total output of the PID controller into a control instruction to control the posture of the underwater robot.

[0050] In the embodiment of the present invention, the PID controller parameters are set, including the proportional coefficient Determines the degree of direct response of the controller to the error. The larger the proportional coefficient, the stronger the controller's response to the error. Used to eliminate the steady-state error of the system. The integral term will continue to accumulate errors until the error is zero. Predict the change trend of the error to adjust the control output in advance to reduce overshoot and oscillation. Initialize the PID controller and initialize the integral term to 0 to avoid initial integral shock. Initialize the last error to 0 as a benchmark for calculating subsequent error changes. Read the preset expected attitude angle, which is the target attitude that the underwater robot should achieve. Read the current attitude angle after fusion, which is the actual attitude of the underwater robot.

[0051] The attitude error is obtained by the difference between the expected attitude angle and the current attitude angle, that is, ;in, is the attitude error; is the desired posture angle; is the current posture angle; Multiply the attitude error by the proportional coefficient to obtain the proportional control component, that is, ;in, For proportional control of the component; is the proportional coefficient; the attitude error is added to the integral term and multiplied by the integral coefficient to obtain the integral control component, that is, ;in, is the integral control component; is the integral coefficient; calculate the difference between the current error and the previous error, multiply it by the differential coefficient, and get the differential control component, that is, ;in, is the differential control component; is the differential coefficient; is the attitude error Small changes in For time The total output of the PID controller is obtained by adding the proportional, integral and differential control components, that is, ;in, is the total output of the controller; the total output of the PID controller is converted into specific control instructions, such as the speed of the propeller, the angle of the servo, etc. These control instructions will be sent to the actuator of the underwater robot to adjust the robot's posture.

[0052] The PID controller can accurately control the posture of the underwater robot so that it can quickly and accurately reach the desired posture angle. The introduction of the differential control component can predict the trend of the error change and adjust the control output in advance, thereby reducing the overshoot and oscillation of the system. The role of the integral control component is to eliminate the steady-state error of the system and ensure that the underwater robot can stably maintain the desired posture angle for a long time. By adjusting the parameters of the PID controller ( , , ), which can flexibly adapt to different control requirements and system characteristics. In a preferred embodiment of the present invention, the above step 6, maintaining the wall-adhering posture by using a PID controller according to the fused data to achieve wall-adhering walking and performing the operation, may include: According to the task requirements, an ideal wall-sticking posture angle is pre-set. This angle is the posture that the underwater robot should maintain when performing the wall-sticking walking task. The calculated posture error is used as the input of the PID controller. According to the posture error, the PID controller reduces the posture error through the calculated control output, so that the posture of the underwater robot approaches the preset wall-sticking posture. The output of the PID controller is converted into specific control instructions. These instructions include adjusting the speed of the propeller, changing the angle of the servo, etc., to achieve precise control of the underwater robot's posture. After receiving these instructions, the control system of the underwater robot will adjust the corresponding actuators (such as propellers, servos, etc.) to change the robot's posture and motion trajectory.

[0053] Through continuous monitoring and adjustment, the underwater robot can maintain a stable wall-attached posture. While maintaining the wall-attached posture, the underwater robot can perform various tasks, such as cleaning, inspection, maintenance, etc. These tasks usually require the robot to be stably close to the wall in a specific posture.

[0054] like Figure 2 As shown, an embodiment of the present invention further provides a system information processing system 20 based on power data, comprising: The acquisition module 21 is used to acquire multiple sets of altimeter data and inertial navigation data, and process the altimeter data and inertial navigation data to obtain optimized altimeter data; compare the differences between the optimized multiple sets of altimeter data, and when the difference is less than a threshold, perform a wall-attaching preparation operation; and calculate the roll angle and pitch angle of the robot using the altimeter data difference and the robot's geometric parameters; The processing module 22 is used to fuse the angles calculated by the altimeter data and the inertial navigation data to obtain a fused attitude angle; determine the attitude error according to the difference between the expected attitude angle and the fused attitude angle, and use the PID controller to calculate the control output to control the attitude of the underwater robot; maintain the wall-adhering attitude according to the fused data and use the PID controller to achieve wall-adhering walking and perform operations.

[0055] It should be noted that the system is a system corresponding to the above method, and all implementation methods in the above method embodiment are applicable to this embodiment and can achieve the same technical effect.

[0056] The embodiment of the present invention further provides a computing device, comprising: a processor, a memory storing a computer program, wherein when the computer program is executed by the processor, the method described above is executed. All implementations in the above method embodiment are applicable to this embodiment and can achieve the same technical effect.

[0057] The embodiment of the present invention also provides a computer-readable storage medium storing instructions, which, when executed on a computer, enable the computer to execute the method described above. All implementations in the above method embodiment are applicable to this embodiment and can achieve the same technical effect.

[0058] The above embodiments are only preferred technical solutions of the present invention and should not be regarded as limiting the present invention. The protection scope of the present invention shall be the technical solutions recorded in the claims, including equivalent replacement solutions of the technical features in the technical solutions recorded in the claims. That is, equivalent replacement improvements within this scope are also within the protection scope of the present invention.

Claims

1. A method for assisting adjustment of the posture of an underwater robot walking against a wall, characterized in that The following steps are involved: Acquire multiple sets of altimeter data and inertial navigation data, and process the altimeter data and inertial navigation data to obtain optimized altimeter data; Compare the differences between the optimized multiple groups of altimeter data, and when the difference is less than a threshold, perform a wall-attaching preparation operation; Using the difference in altimeter data and the robot's geometric parameters, the robot's roll and pitch angles are calculated; Fuse the angles calculated by the altimeter data and the inertial navigation data to obtain the fused attitude angle; Determine the attitude error according to the difference between the desired attitude angle and the fused attitude angle, and use the PID controller to calculate the control output to control the attitude of the underwater robot; The fused data is used to maintain the wall-adhering posture using a PID controller to achieve wall-adhering walking and perform tasks.

2. The method for assisting adjustment of the underwater robot's wall-walking posture according to claim 1, characterized in that: The method of acquiring multiple sets of altimeter data and inertial navigation data, and processing the altimeter data and inertial navigation data to obtain optimized altimeter data includes the following steps: Acquire multiple sets of altitude data through the altimeter sensor, and acquire acceleration and angular velocity data through the inertial navigation system; Filtering the altimeter data to obtain smoothed altimeter data, and calibrating the inertial navigation data to obtain calibrated acceleration and angular velocity data; Establish the state equation based on the inertial navigation data, and use the altimeter data as the observation value to establish the observation equation; According to the altimeter data and inertial navigation data, the initial state estimate is set, and the prediction error covariance matrix and the process and observation noise covariance matrix are initialized; Use the state estimate and state equation to predict the current state and update the prediction error covariance matrix. When new altimeter data arrives, calculate the Kalman gain. Update the state estimate using the Kalman gain and new observations to fuse the prediction with the observation information and update the prediction error covariance matrix; Continuously iterate and update the status every time new altimeter data arrives; When the preset number of iterations is reached, the filter outputs the optimized height data.

3. The method for assisting adjustment of the underwater robot's wall-walking posture according to claim 2, characterized in that: The equation of state is calculated as: ； in, represents the state vector at time k; represents the state transfer matrix; Indicates at the last time point The state vector of represents the control input matrix; Indicated in Control input at all times; represents process noise.

4. The method for assisting adjustment of the underwater robot's wall-walking posture according to claim 2, characterized in that: The calculation formula of the observation equation is: ； in, Indicates at time Observed value of represents the observation matrix; represents the observation noise.

5. The method for assisting adjustment of the underwater robot's wall-walking posture according to claim 2, characterized in that: The calculation formula of Kalman gain is: ； in, Indicates at time The Kalman gain of represents the prediction error covariance matrix; represents the transpose of the observation matrix; represents the observation noise covariance matrix.

6. The method for assisting adjustment of the underwater robot's wall-walking posture according to claim 2, characterized in that: The calculation formula of the fused posture angle is: ； in, Indicates the posture angle after fusion; Represents the attitude angle calculated based on the difference in altimeter data; Indicates the attitude angle calculated based on the inertial navigation data; Represents the dynamically adjusted fusion weight.

7. The method for assisting adjustment of the underwater robot's wall-walking posture according to claim 1, characterized in that: The method of determining the attitude error according to the difference between the expected attitude angle and the fused attitude angle, and using a PID controller to calculate the control output to control the attitude of the underwater robot includes the following steps: Set the parameters of the PID controller, including the proportional coefficient, integral coefficient and differential coefficient; Initialize the integral term and the last error of the PID controller to 0, and obtain the preset desired attitude angle and the fused current attitude angle to obtain the attitude error; According to the attitude error, the proportional, integral and differential control components are calculated by the PID controller to obtain the total output of the PID controller; The total output of the PID controller is converted into control instructions to control the posture of the underwater robot.

8. An auxiliary adjustment system for an underwater robot's wall-climbing walking posture, characterized in that: A method for assisting adjustment of the wall-walking posture of an underwater robot according to any one of claims 1 to 7 is adopted, comprising: The acquisition module is used to acquire multiple sets of altimeter data and inertial navigation data, and process the altimeter data and inertial navigation data to obtain optimized altimeter data; compare the differences between the optimized multiple sets of altimeter data, and when the difference is less than a threshold, perform a wall-attaching preparation operation; and calculate the roll angle and pitch angle of the robot using the altimeter data difference and the robot's geometric parameters; The processing module is used to fuse the angles calculated by the altimeter data and the inertial navigation data to obtain the fused attitude angle; determine the attitude error according to the difference between the expected attitude angle and the fused attitude angle, and use the PID controller to calculate the control output to control the attitude of the underwater robot; maintain the wall-adhering attitude according to the fused data and use the PID controller to achieve wall-adhering walking and perform operations.

9. A computing device, characterized in that include: one or more processors; A storage device for storing one or more programs. When the one or more programs are executed by the one or more processors, the one or more processors implement an auxiliary adjustment method for the wall-walking posture of an underwater robot as described in any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a program, which, when executed by a processor, implements an auxiliary adjustment method for the wall-walking posture of an underwater robot as described in any one of claims 1 to 7.

Images