Running attitude control method and system for heavy-load robot and storage medium

Abstract

The invention discloses an operation attitude control method and system for a heavy-load robot and a storage medium, and relates to the technical field of robot control, a collection feature set E is constructed in real time, and load distribution information and a gravity center offset G obtained by a monitoring system in the robot are combined to obtain a control feature set E; and the operation posture and the load condition of the robot in a complex environment are effectively evaluated. The robot can accurately adjust the advancing posture and dynamically generate a control command by executing the vector W, and real-time control and correction are achieved. By continuously tracking and counting the variable quantity Z of the gravity center position of the robot, adjustment and early warning are triggered in time, the stability and safety of the robot in complex tasks are further improved, and compared with a traditional method, the method can respond to dynamic factors such as ground fluctuation and load change in real time, and the real-time performance of the robot is improved. And the adaptability to the external environment and the state change of the robot is enhanced. Especially in the operation process, an accurate attitude adjustment and gravity center shift evaluation mechanism is used.

Description

Technical Field

[0001] The present invention relates to the field of robot control technology, and in particular to an operation posture control method, system and storage medium for a heavy-load robot. Background Art

[0002] With the advancement of science and technology, automation technology has gradually been applied in many fields, especially in heavy-load operations such as industry, logistics, and agriculture. As an important part of intelligent equipment, heavy-load robots are playing an increasingly important role. This type of robot is mainly used for tasks such as handling, loading and unloading, and transportation, and can replace manual labor for efficient and precise operations. Specifically, heavy-load robots can not only handle common handling operations, such as bulk cargo handling in warehouses, but are also widely used in special scenarios, such as dangerous goods transportation, construction materials handling, port loading and unloading, and other fields.

[0003] In the invention of Chinese patent application No. CN202411359674.0, an intelligent control method, device and equipment of a heavy-load robot are disclosed, which belongs to the field of heavy-load robots. The present invention calculates the torque information of each lever in each operable posture according to the mass information of the goods and the operable posture data set of the heavy-load robot, and finally obtains the operating posture of the heavy-load robot in the process of picking and unloading according to the torque information of each lever in the operable posture and the motion path of the position of the goods during the handling process, and controls the heavy-load robot according to the operating posture of the heavy-load robot in the process of picking and unloading. The present invention fully considers the change of the maximum load torque of each lever of the heavy-load robot under different operating postures, so that the heavy-load robot changes the operating posture according to the maximum load torque characteristics, which can increase the service life of the heavy-load robot, reduce the maintenance cost of the heavy-load robot, and make the control of the heavy-load robot more reasonable.

[0004] It can be seen that the above invention can solve limited scenarios and situations. For example, most of the existing control methods can only preset load distribution and posture adjustment strategies, but cannot be adjusted according to real-time feedback in actual operation. This results in the robot being unable to respond in time when encountering sudden ground undulations or unevenness during transportation, thereby affecting the overall transportation efficiency and task completion quality. Especially in complex environments, the impact of ground unevenness on robots is very significant. At the same time, most of the existing heavy-load robot posture control methods are based on predetermined load distribution and ground conditions for control design. These methods usually assume that the ground conditions encountered by the robot during operation are relatively flat or predictable. However, in practical applications, especially in industrial environments, the ground often has irregular undulations, potholes and obstacles. These uneven grounds directly affect the load distribution and stability of the robot. The existing technology is often unable to monitor and compensate for the impact of ground unevenness in real time. Once the robot encounters uneven ground, uneven load or center of gravity offset may occur, resulting in unstable robot posture, and even abnormal situations such as tilting, tipping over or insufficient power. The existing control methods rely more on static load calculation and ground condition prediction, and lack dynamic and real-time adaptability. Summary of the invention

[0005] In view of the deficiencies in the prior art, the present invention provides a method, system and storage medium for controlling the running posture of a heavy-load robot, which solves the problems mentioned in the background technology.

[0006] To achieve the above objectives, the present invention is implemented through the following technical solutions: A running posture control method for a heavy-load robot comprises the following steps: S1, by integrating a laser scanner and a depth camera sensor on the robot, real-time data of the ground of the current robot's forward path is obtained in real time, and after collection and analysis, it is recorded as a collection feature set E; S2. Obtain the load distribution of the robot by establishing a monitoring system integrated with the robot, and calculate the center of gravity offset △G of the current robot according to the load state, and output the state evaluation vector B of the robot; S3, by analyzing the collected feature set E and the state evaluation vector B, the robot's current forward path ground, the impact on the robot's posture and load, and obtaining the robot's adaptive posture adjustment vector △D; S4, the robot adjusts the current forward posture of the robot based on the acquired posture adjustment vector △D, obtains the adjusted execution vector W, and generates a control command for the execution vector W to dynamically adjust the forward posture of the robot and mark the center of gravity position G of the current robot; S5. By counting the change △Z of the robot's center of gravity position within a fixed period and comparing it with the preset robot operation center of gravity offset threshold Zthe, the current robot's operation center of gravity offset evaluation result is obtained, and iterative optimization is performed based on the operation center of gravity offset evaluation result.

[0007] Preferably, said S1 includes S11 and S12; S11. By integrating a laser scanner and a depth camera sensor on the robot, real-time data of the ground of the current robot's forward path is obtained in real time, and the height value Zg(x, y) at the position (x, y), the slope change value Gd(x, y) at the position (x, y), and the obstacle mark ZAbs at the position (x, y) are obtained to form a ground feature data set G; Among them, the height value Zg (x, y) is obtained by matching and analyzing the laser scanning position height information Zc (x, y) and the depth camera sensor recognition position height Zl (x, y); S12. Marking the ground feature data set G with timestamp information and unique identification information of the load, thereby forming a collection feature set E of the ground of the forward path of the robot transporting the load at time t.

[0008] Preferably, the laser scanning position height information Zc (x, y) is obtained by real-time measurement of the laser scanning height information by the laser scanner, and the depth camera sensor recognition position height Zl (x, y) is obtained by real-time analysis of the depth camera sensor; The height value Zg (x, y) is obtained by the following calculation formula: ; Wherein, max represents the maximum value function, and max(Zc(x, y), Zl(x, y)) represents the maximum height value of the laser scanning position height information Zc(x, y) and the depth camera sensor recognition position height Zl(x, y) at the same position (x, y); The slope change value Gd(x, y) is obtained by the following calculation formula: ; Wherein, arctan represents the inverse tangent function, and dx represents the horizontal distance between the height values ​​at the position (x, y) and the position (x+1, y); The obstacle identification ZAbs is obtained by comparing with a preset ground height threshold Zgthe of the robot's forward path; The comparison is made in the following ways: ; When the obstacle mark ZAbs ≥ the ground height threshold Zgthe of the robot's forward path, the height value of the robot's forward path ground at the position (x, y) is judged to be abnormal, and the obstacle mark ZAbs = 1; When the obstacle mark ZAbs is less than the ground height threshold Zgthe of the robot's forward path, it is determined that the height value of the ground of the robot's forward path at the position (x, y) is normal, and the obstacle mark ZAbs=0.

[0009] Preferably, S2 includes S21 and S22; S21. Obtain the load state of the robot by establishing a monitoring system integrated with the robot, including using an accelerometer and a force sensor to collaboratively obtain the load mass of each component of the robot, to obtain the center of gravity position G of the robot in the current operating state, and compare it with the preset target center of gravity position Gtarget of the robot to obtain the current center of gravity offset △G of the robot; S22. Integrate the center of gravity offset △G with the robot's running posture angle to output the robot's state evaluation vector B, where the running posture angle includes the pitch angle Pitch, the roll angle Roll and the heading angle Yaw, which are specifically obtained through the gyroscope sensor, the acceleration sensor and the magnetometer sensor.

[0010] Preferably, the center of gravity position G is obtained by the following calculation formula: ; Where Xg, Yg and Zg represent the horizontal, vertical and vertical coordinates of the total center of gravity of the robot, n represents the total number of robot components, q(i) represents the mass of the i-th component, x(i), y(i) and z(i) represent the horizontal, vertical and vertical coordinates of the i-th component, respectively. The center of gravity offset ΔG is The calculation formula is obtained, wherein the target center of gravity position Gtarget includes the target horizontal axis Xtarget, the target vertical axis Ytarget and the target longitudinal axis Ztarget of the target center of gravity position.

[0011] Preferably, said S3 includes S31; S31, by analyzing the influence of the robot's forward path ground on the robot's posture and load based on the collected feature set E and the state evaluation vector B, calculate and obtain the robot's adaptive posture adjustment vector △D, including the horizontal axis position offset △X, the vertical axis position offset △Y, the longitudinal axis position offset △Z, the pitch angle adjustment △Pitch, the roll angle adjustment △Roll and the heading angle adjustment △Yaw; The horizontal axis position offset △X is obtained by the following calculation formula: ; Wherein, k1 represents the first proportional coefficient, Zg(x, y) represents the height value at the position (x, y), and Gd(x, y) represents the slope change value at the position (x, y); The vertical axis position offset △Y is obtained by the following calculation formula: ; Wherein, k2 represents the second proportional coefficient, which is specifically used to adjust the influence of the height difference of the path on the vertical axis position offset △Y axis offset; The longitudinal axis position offset ΔZ is obtained by the following calculation formula: ; Wherein, k3 represents the third proportional coefficient, which is specifically used to control the influence of the obstacle on the longitudinal axis position offset △Z, and ZAbs represents the obstacle identifier; The pitch angle adjustment amount ΔPitch is obtained by the following calculation formula: ; Wherein, k4 and k7 represent the fourth proportional coefficient and the seventh proportional coefficient respectively, which are specifically used to adjust the influence of the offset on the pitch angle adjustment amount △Pitch; The roll angle adjustment amount ΔRoll is obtained by the following calculation formula: ; Wherein, k5 and k8 represent the fifth proportional coefficient and the eighth proportional coefficient respectively, which are specifically used to adjust the influence of the offset on the roll angle adjustment amount △Roll; The heading angle adjustment ΔYaw is obtained by the following calculation formula: ; Wherein, k6 and k9 represent the sixth proportional coefficient and the ninth proportional coefficient respectively, which are specifically used to adjust the influence of the offset on the heading angle adjustment amount △Yaw.

[0012] Preferably, the S4 includes S41; S41, the robot adjusts the current forward posture of the robot based on the acquired posture adjustment vector △D, specifically by converting the horizontal axis position offset △X, the vertical axis position offset △Y, the longitudinal axis position offset △Z, the pitch angle adjustment △Pitch, the roll angle adjustment △Roll and the heading angle adjustment △Yaw in the posture adjustment vector △D into a control command, and the robot's internal motion control system acts on the mechanical parts to change the robot's forward posture, and the adjusted execution vector W is obtained, and the execution vector W is generated. The control command is used to dynamically adjust the robot's forward posture and mark the current robot's center of gravity position G; The execution vector W is obtained by the following conversion formula: ; Where f represents the transfer function preset by the control system.

[0013] Preferably, the S5 includes S51; S51, by counting the fluctuation change information of the center of gravity position G of the robot within a fixed period, the center of gravity position change △Z is obtained, and compared with the preset robot operation center of gravity offset threshold Zthe, the current robot operation center of gravity offset evaluation result is obtained, and an iterative optimization mechanism is performed according to the operation center of gravity offset evaluation result; The center of gravity position change ΔZ is obtained by the following calculation formula: ; Where M represents the total number of sampling times within a fixed period, t(j+1, G) represents the centroid position at the j+1th sampling time t, and t(j, G) represents the centroid position at the jth sampling time t; The running center of gravity offset evaluation result is obtained by the following comparison method: When the center of gravity position change △Z ≥ the robot operation center of gravity offset threshold Zthe, the operation center of gravity offset evaluation result is obtained as an abnormal result, triggering the iterative optimization mechanism, including adjusting the acquisition of the posture adjustment vector △D, correcting the center of gravity position and the posture angle, specifically by adjusting the first proportional coefficient k1 to the ninth proportional coefficient k9; When the center of gravity position change △Z is less than the robot operation center of gravity offset threshold Zthe, the operation center of gravity offset evaluation result is obtained as a normal result.

[0014] A running posture control system for a heavy-load robot, comprising a forward path acquisition module, a machine state acquisition module, a posture analysis and adjustment module, an adjustment generation and marking module, and an iterative optimization module; The forward path acquisition module acquires the real-time data of the ground of the current forward path of the robot in real time by integrating a laser scanner and a depth camera sensor on the robot, and records it as a collection feature set E after collection and analysis; The machine state acquisition module acquires the load distribution of the robot by establishing a monitoring system integrated with the robot, and calculates the center of gravity offset △G of the current robot according to the load state, and outputs the state evaluation vector B of the robot; The posture analysis and adjustment module calculates and obtains the posture adjustment vector △D of the robot for adaptive adjustment by analyzing the influence of the ground surface on the robot's forward path and the robot's load through the collected feature set E and the state evaluation vector B; The adjustment generation and marking module adjusts the forward posture of the current robot based on the acquired posture adjustment vector △D through the robot, obtains the adjusted execution vector W, and generates a control command for the execution vector W to dynamically adjust the forward posture of the robot and mark the center of gravity position G of the current robot; The iterative optimization module obtains the current robot's operating center of gravity offset evaluation result by counting the robot's center of gravity position change △Z within a fixed period and comparing it with the preset robot operating center of gravity offset threshold Zthe, and performs iterative optimization based on the operating center of gravity offset evaluation result.

[0015] A running posture control storage medium for a heavy-load robot stores a computer program. When the computer program is executed, a running posture control method for the heavy-load robot is implemented.

[0016] The present invention provides a method, system and storage medium for controlling the running posture of a heavy-load robot, which have the following beneficial effects: (1) The collected feature set E is constructed in real time, and combined with the load distribution information and center of gravity offset △G obtained by the robot's internal monitoring system, the robot's running posture and load conditions in complex environments are effectively evaluated. The robot can accurately adjust its forward posture and dynamically generate control commands by executing vector W to achieve real-time control and correction. By continuously tracking and counting the robot's center of gravity position change △Z, timely triggering adjustments and warnings, the robot's stability and safety in complex tasks are further improved. Compared with traditional methods, this method can not only respond to dynamic factors such as ground undulations and load changes in real time, but also enhance its adaptability to changes in the external environment and the robot's own state. In particular, during operation, through precise posture adjustment and center of gravity offset evaluation mechanisms, the traditional method effectively solves the problems of being unable to adjust the robot's posture and load distribution in real time and ignoring the impact of ground unevenness and environmental changes on the robot's running stability.

[0017] (2) By calculating and optimizing the current center of gravity position G of the robot and comparing it with the target center of gravity position Gtarget, the center of gravity offset △G is obtained, which effectively guides the robot to maintain balance on unstable ground. For example, in response to factors such as path height changes, slope differences, and obstacle positions, the robot can reasonably adjust its posture angle to maintain stability. By using different proportional coefficients, flexible responses to factors such as path changes, load offsets, and posture adjustments are achieved, effectively avoiding over-adjustment or under-adjustment and ensuring the smooth operation of the robot. This method not only improves the robot's ability to adapt to complex environments, but also greatly improves its operating stability on dynamic and changing ground, further optimizing the robot's posture control system.

[0018] (3) By precisely controlling the robot's posture adjustment vector △D, the robot can always adapt to changes in the ground and load in real time during its forward movement. By statistically analyzing the fluctuations in the robot's center of gravity position within a fixed period, the center of gravity position change △Z is obtained and compared with the preset center of gravity offset threshold Zthe to evaluate the robot's operating center of gravity offset. When it is detected that the center of gravity position change △Z exceeds the preset threshold, the system triggers an iterative optimization mechanism to further adjust the posture adjustment vector △D, and optimize the robot's forward path by correcting the center of gravity position and posture angle. This mechanism can respond to abnormal conditions in a timely manner, ensuring that the robot continues to operate stably under heavy loads and complex environments, and effectively avoiding posture instability problems caused by load changes or path mutations. Through dynamic adjustment and optimization, the robot's adaptability and operational efficiency in various complex terrains are improved. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] Figure 1 A schematic diagram of the steps of a method for controlling the running posture of a heavy-load robot according to the present invention; Figure 2 The present invention is a schematic diagram of a block diagram of an operation posture control system for a heavy-load robot. DETAILED DESCRIPTION

[0020] The technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, rather than all the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of the present invention. Example 1

[0021] The present invention provides a method for controlling the running posture of a heavy-load robot. Figure 1 , including the following steps: S1, by integrating a laser scanner and a depth camera sensor on the robot, real-time data of the ground of the current robot's forward path is acquired in real time, and after collection and analysis, it is recorded as a collection feature set E; S2. Obtain the load distribution of the robot by establishing a monitoring system integrated with the robot, and calculate the center of gravity offset △G of the current robot according to the load state, and output the state evaluation vector B of the robot; S3, by analyzing the collected feature set E and the state evaluation vector B, the robot's current forward path ground, the impact on the robot's posture and load, and obtaining the robot's adaptive posture adjustment vector △D; S4, the robot adjusts the current forward posture of the robot based on the acquired posture adjustment vector △D, obtains the adjusted execution vector W, and generates a control command for the execution vector W to dynamically adjust the forward posture of the robot and mark the center of gravity position G of the current robot; S5. By counting the change △Z of the robot's center of gravity position within a fixed period and comparing it with the preset robot operation center of gravity offset threshold Zthe, the current robot's operation center of gravity offset evaluation result is obtained, and iterative optimization is performed based on the operation center of gravity offset evaluation result.

[0022] In this embodiment, through steps S1 to S5, ground data is collected and analyzed in real time, and a collection feature set E is constructed. In combination with the load distribution information and the center of gravity offset △G obtained by the robot's internal monitoring system, the robot's running posture and load conditions in a complex environment are effectively evaluated. The robot can accurately adjust its forward posture and dynamically generate control commands by executing vector W to achieve real-time control and correction. By continuously tracking and counting the robot's center of gravity position change △Z, and comparing it with the preset center of gravity offset threshold Zthe, this method can automatically evaluate and optimize the robot's running state within a fixed period, trigger adjustments and warnings in a timely manner, and further improve the robot's stability and safety in complex tasks. Compared with traditional methods, this method can not only respond to dynamic factors such as ground undulations and load changes in real time, but also enhance the adaptability to changes in the external environment and the robot's own state through the close association between the collection feature set E and the state evaluation vector B. In particular, during operation, through precise posture adjustment and center of gravity offset evaluation mechanism, the traditional method can effectively solve the problems of being unable to adjust the robot's posture and load distribution in real time, ignoring the impact of ground unevenness and environmental changes on the robot's running stability, etc. This innovative approach ensures the continuous operation of the robot in unstable environments, reduces maintenance costs and potential failure risks, and further improves the intelligence level of heavy-load robots. Example 2

[0023] This embodiment is explained in Example 1, please refer to Figure 1 , specifically: the S1 includes S11 and S12; S11. By integrating a laser scanner and a depth camera sensor on the robot, real-time data of the ground of the current robot's forward path is obtained in real time, and the height value Zg(x, y) at the position (x, y), the slope change value Gd(x, y) at the position (x, y), and the obstacle mark ZAbs at the position (x, y) are obtained to form a ground feature data set G; Among them, the height value Zg (x, y) is obtained by matching and analyzing the laser scanning position height information Zc (x, y) and the depth camera sensor recognition position height Zl (x, y); S12. Marking the ground feature data set G with timestamp information and unique identification information of the load, thereby forming a collection feature set E of the ground of the forward path of the robot transporting the load at time t.

[0024] The laser scanning position height information Zc (x, y) is obtained by the laser scanning height information measured in real time by the laser scanner, and the depth camera sensor recognition position height Zl (x, y) is obtained by real-time analysis of the depth camera sensor; The height value Zg (x, y) is obtained by the following calculation formula: ; In the formula, max represents the maximum value function, and max(Zc(x, y), Zl(x, y)) represents the maximum height value of the laser scanning position height information Zc(x, y) and the depth camera sensor recognition position height Zl(x, y) at the same position (x, y). In this way, the maximum value at the forward position (x, y) can be continuously obtained as the robot's forward path continues to change; The slope change value Gd(x, y) is obtained by the following calculation formula: ; Wherein, arctan represents the inverse tangent function, and dx represents the horizontal distance between the height values ​​at the position (x, y) and the position (x+1, y); The obstacle identification ZAbs is obtained by comparing with a preset ground height threshold Zgthe of the robot's forward path; The comparison is made in the following ways: ; When the obstacle mark ZAbs ≥ the ground height threshold Zgthe of the robot's forward path, the height value of the robot's forward path ground at the position (x, y) is judged to be abnormal, and the obstacle mark ZAbs = 1; When the obstacle mark ZAbs is less than the ground height threshold Zgthe of the robot's forward path, it is determined that the height value of the ground of the robot's forward path at the position (x, y) is normal, and the obstacle mark ZAbs=0.

[0025] In this embodiment, the height value Zg(x, y), slope change value Gd(x, y) and obstacle identification ZAbs at different positions (x, y) are accurately obtained through the real-time acquisition of the ground data of the forward path, and the ground feature data set G on the forward path is continuously updated by means of maximum value selection and difference calculation. These data provide the robot with comprehensive ground state information, enabling the robot to perceive the changes in the forward path in real time, including ground undulations, slope changes and potential obstacles. In this process, by marking the ground feature data at each moment with timestamps and unique identification information of the carrier, a complete collection feature set E is finally formed, providing accurate environmental feedback for subsequent robot posture adjustment and load control. By comprehensively using the three parameters of height value Zg(x, y), slope change value Gd(x, y) and obstacle identification ZAbs, the real-time adaptability of the robot to the environment is ensured. By dynamically updating these ground feature data, the robot can accurately judge the ground state of the forward path, identify ground obstacles in time and take corresponding countermeasures. In addition, the maximum value selection mechanism of the height value Zg(x, y) improves the accuracy of the data, and the calculation of the slope change value Gd(x, y) can effectively help the robot identify the change in the road slope and avoid losing balance on uneven ground. Finally, by judging the obstacle mark ZAbs, the robot can avoid the impact of obstacles on the forward route, thereby ensuring the stable operation of the robot in complex and dynamic environments. This innovative method significantly improves the robot's perception and adaptability under different ground conditions, laying a solid foundation for subsequent posture adjustment and load control. Example 3

[0026] This embodiment is explained in Example 2. Please refer to Figure 1 , specifically: S2 includes S21 and S22; S21. Obtain the load state of the robot by establishing a monitoring system integrated with the robot, including using an accelerometer and a force sensor to collaboratively obtain the load mass of each component of the robot, to obtain the center of gravity position G of the robot in the current operating state, and compare it with the preset target center of gravity position Gtarget of the robot to obtain the current center of gravity offset △G of the robot; S22. Integrate the center of gravity offset △G with the robot's running posture angle to output the robot's state evaluation vector B, where the running posture angle includes the pitch angle Pitch, the roll angle Roll and the heading angle Yaw, which are specifically obtained through the gyroscope sensor, the acceleration sensor and the magnetometer sensor.

[0027] The center of gravity position G is obtained by the following calculation formula: ; Where Xg, Yg and Zg represent the horizontal, vertical and vertical coordinates of the total center of gravity of the robot, n represents the total number of robot components, q(i) represents the mass of the i-th component, x(i), y(i) and z(i) represent the horizontal, vertical and vertical coordinates of the i-th component, respectively. The center of gravity offset ΔG is The calculation formula is obtained, wherein the target center of gravity position Gtarget includes the target horizontal axis Xtarget, the target vertical axis Ytarget and the target longitudinal axis Ztarget of the target center of gravity position.

[0028] The S3 includes S31; S31, by analyzing the influence of the robot's forward path ground on the robot's posture and load based on the collected feature set E and the state evaluation vector B, calculate and obtain the robot's adaptive posture adjustment vector △D, including the horizontal axis position offset △X, the vertical axis position offset △Y, the longitudinal axis position offset △Z, the pitch angle adjustment △Pitch, the roll angle adjustment △Roll and the heading angle adjustment △Yaw; The horizontal axis position offset △X is obtained by the following calculation formula: ; In the formula, k1 represents the first proportional coefficient, which is specifically used to control the influence of the change in ground height on the horizontal axis position offset △X, Zg(x, y) represents the height value at the position (x, y), (Zg(x, y)-Zg(x-1, y) represents the height difference of the current path position, reflecting the change of the front and rear ground, Gd(x, y) represents the slope change value at the position (x, y), which is used to indicate the slope difference between the current position and the adjacent position on the path. This parameter is used to adjust the adjustment that the robot needs to make in the forward direction; The vertical axis position offset △Y is obtained by the following calculation formula: ; Wherein, k2 represents the second proportional coefficient, which is specifically used to adjust the influence of the height difference of the path on the vertical axis position offset △Y axis offset; The longitudinal axis position offset ΔZ is obtained by the following calculation formula: ; Wherein, k3 represents the third proportional coefficient, which is specifically used to control the influence of the obstacle on the longitudinal axis position offset △Z, and ZAbs represents the obstacle identifier; The first proportional coefficient k1, the second proportional coefficient k2 and the third proportional coefficient k3 are related to the robot's center of gravity offset (for example, the adjustment amount of the position error), and are usually in the range of [0.1, 1]. The specific values ​​are set by the user; The pitch angle adjustment amount ΔPitch is obtained by the following calculation formula: ; Wherein, k4 and k7 represent the fourth proportional coefficient and the seventh proportional coefficient respectively, which are specifically used to adjust the influence of the offset on the pitch angle adjustment amount △Pitch; The roll angle adjustment amount ΔRoll is obtained by the following calculation formula: ; Wherein, k5 and k8 represent the fifth proportional coefficient and the eighth proportional coefficient respectively, which are specifically used to adjust the influence of the offset on the roll angle adjustment amount △Roll; The heading angle adjustment ΔYaw is obtained by the following calculation formula: ; Wherein, k6 and k9 represent the sixth proportional coefficient and the ninth proportional coefficient respectively, which are specifically used to adjust the influence of the offset on the heading angle adjustment amount △Yaw; Among them, the fourth proportional coefficient k4, the fifth proportional coefficient k5 and the sixth proportional coefficient k6 are related to the robot's attitude angle (pitch, roll, heading) and path characteristics (such as slope, obstacles, undulations, etc.), usually in the range of [0.1, 1], and the specific values ​​are set by the user; The seventh proportional coefficient k7, the eighth proportional coefficient k8 and the ninth proportional coefficient k9 are related to the response sensitivity of the load, path and posture changes, and are usually in the range of [0.05, 0.5]. The specific values ​​are set by the user; For example, when the path has large longitudinal undulations, the longitudinal axis position offset △Z has a greater impact on the pitch angle adjustment △Pitch, so the fourth proportional coefficient k4 can be set to 0.8, while the adjustment of the transverse axis position offset △X can be set to a lower coefficient, the seventh proportional coefficient k7=0.3, to avoid excessive adjustment.

[0029] In this embodiment, by integrating accelerometers, force sensors, gyroscopes, acceleration sensors and magnetometer sensors, the robot's load status, center of gravity position and posture angle can be accurately obtained, and then the robot's forward posture can be dynamically calculated and adjusted. In particular, by obtaining the load mass of each component, the robot's current center of gravity position G is calculated and optimized, and compared with the target center of gravity position Gtarget, the center of gravity offset △G is obtained, which effectively guides the robot to maintain balance on unstable ground. For example, in response to factors such as path height changes, slope differences and obstacle positions, the robot can reasonably adjust its posture angle to maintain stability. By using different proportional coefficients, flexible responses to factors such as path changes, load offsets, and posture adjustments are achieved, effectively avoiding over-adjustment or under-adjustment, and ensuring the smooth operation of the robot. This method not only improves the robot's adaptability to complex environments, but also greatly improves its operating stability on dynamic and changing ground, and further optimizes the robot's posture control system. Example 4

[0030] This embodiment is explained in Example 3, please refer to Figure 1 Specifically: S4 includes S41; S41, the robot adjusts the current forward posture of the robot based on the acquired posture adjustment vector △D, specifically by converting the horizontal axis position offset △X, the vertical axis position offset △Y, the longitudinal axis position offset △Z, the pitch angle adjustment △Pitch, the roll angle adjustment △Roll and the heading angle adjustment △Yaw in the posture adjustment vector △D into a control command, and the robot's internal motion control system acts on the mechanical parts to change the robot's forward posture, and the adjusted execution vector W is obtained, and the execution vector W is generated. The control command is used to dynamically adjust the robot's forward posture and mark the current robot's center of gravity position G; The execution vector W is obtained by the following conversion formula: ; Wherein, f represents the conversion function preset by the control system, which is specifically used to convert the vector in the attitude adjustment vector △D into a specific control signal.

[0031] The S5 includes S51; S51, by counting the fluctuation change information of the center of gravity position G of the robot within a fixed period, the center of gravity position change △Z is obtained, and compared with the preset robot operation center of gravity offset threshold Zthe, the current robot operation center of gravity offset evaluation result is obtained, and an iterative optimization mechanism is performed according to the operation center of gravity offset evaluation result; The center of gravity position change ΔZ is obtained by the following calculation formula: ; Where M represents the total number of sampling times within a fixed period, t(j+1, G) represents the centroid position at the j+1th sampling time t, and t(j, G) represents the centroid position at the jth sampling time t; The running center of gravity offset evaluation result is obtained by the following comparison method: When the center of gravity position change △Z ≥ the robot operation center of gravity offset threshold Zthe, the operation center of gravity offset evaluation result is obtained as an abnormal result, triggering the iterative optimization mechanism, including adjusting the acquisition of the posture adjustment vector △D, correcting the center of gravity position and the posture angle, specifically by adjusting the first proportional coefficient k1 to the ninth proportional coefficient k9; When the center of gravity position change △Z is less than the robot operation center of gravity offset threshold Zthe, the operation center of gravity offset evaluation result is obtained as a normal result.

[0032] In this embodiment, by precisely controlling the robot's posture adjustment vector △D, it is ensured that the robot can always adapt to the changes of the ground and load in real time during the forward process. By converting each posture adjustment amount in the posture adjustment vector △D into a specific control signal, the mechanical components are adjusted through the motion control system inside the robot, thereby realizing the dynamic adjustment of the robot's forward posture. This flexible adjustment method enables the robot to maintain stable operation in a complex and dynamic environment, improves the adaptability and stability of the system, and in addition, by statistically analyzing the fluctuation changes of the robot's center of gravity position within a fixed period, the center of gravity position change △Z is obtained, and compared with the preset center of gravity offset threshold Zthe, so as to evaluate the robot's running center of gravity offset. When it is detected that the center of gravity position change △Z exceeds the preset threshold, the system will trigger the iterative optimization mechanism, further adjust the posture adjustment vector △D, and optimize the robot's forward path by correcting the center of gravity position and posture angle. This mechanism can respond to abnormal conditions in a timely manner, ensure that the robot continues to operate stably under heavy loads and complex environments, and effectively avoid the problem of posture instability caused by load changes or path mutations. Through dynamic adjustment and optimization, the robot's adaptability and operating efficiency in various complex terrains are improved. Example 5

[0033] A motion control system for heavy-load robots, please refer to Figure 2 ,Specifically: including forward path acquisition module, machine status acquisition module, posture analysis and adjustment module, adjustment generation and marking module and iterative optimization module; The forward path acquisition module acquires the real-time data of the ground of the current forward path of the robot in real time by integrating a laser scanner and a depth camera sensor on the robot, and records it as a collection feature set E after collection and analysis; The machine state acquisition module acquires the load distribution of the robot by establishing a monitoring system integrated with the robot, and calculates the center of gravity offset △G of the current robot according to the load state, and outputs the state evaluation vector B of the robot; The posture analysis and adjustment module calculates and obtains the posture adjustment vector △D of the robot for adaptive adjustment by analyzing the influence of the ground surface on the robot's forward path and the robot's load through the collected feature set E and the state evaluation vector B; The adjustment generation and marking module adjusts the forward posture of the current robot based on the acquired posture adjustment vector △D through the robot, obtains the adjusted execution vector W, and generates a control command for the execution vector W to dynamically adjust the forward posture of the robot and mark the center of gravity position G of the current robot; The iterative optimization module obtains the current robot's operating center of gravity offset evaluation result by counting the robot's center of gravity position change △Z within a fixed period and comparing it with the preset robot operating center of gravity offset threshold Zthe, and performs iterative optimization based on the operating center of gravity offset evaluation result. Example 6

[0034] A running posture control storage medium for a heavy-load robot. Specifically: the storage medium stores a computer program, and when the computer program is executed, a running posture control method for a heavy-load robot is implemented.

[0035] Although embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions and variations may be made to the embodiments without departing from the principles and spirit of the present invention, and that the scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. A method for controlling the running posture of a heavy-load robot, characterized in that: The following steps are involved: S1, by integrating a laser scanner and a depth camera sensor on the robot, real-time data of the ground of the current robot's forward path is obtained in real time, and after collection and analysis, it is recorded as a collection feature set E; S2. Obtain the load distribution of the robot by establishing a monitoring system integrated with the robot, and calculate the center of gravity offset △G of the current robot according to the load state, and output the state evaluation vector B of the robot; S3, by analyzing the collected feature set E and the state evaluation vector B, the robot's current forward path ground, the impact on the robot's posture and load, and obtaining the robot's adaptive posture adjustment vector △D; S4, the robot adjusts the current forward posture of the robot based on the acquired posture adjustment vector △D, obtains the adjusted execution vector W, and generates a control command for the execution vector W to dynamically adjust the forward posture of the robot and mark the center of gravity position G of the current robot; S5. By counting the change △Z of the robot's center of gravity position within a fixed period and comparing it with the preset robot operation center of gravity offset threshold Zthe, the current robot's operation center of gravity offset evaluation result is obtained, and iterative optimization is performed based on the operation center of gravity offset evaluation result.

2. The method for controlling the running posture of a heavy-load robot according to claim 1, characterized in that: Said S1 includes S11 and S12; S11. By integrating a laser scanner and a depth camera sensor on the robot, real-time data of the ground of the current robot's forward path is obtained in real time, and the height value Zg(x, y) at the position (x, y), the slope change value Gd(x, y) at the position (x, y), and the obstacle mark ZAbs at the position (x, y) are obtained to form a ground feature data set G; Among them, the height value Zg (x, y) is obtained by matching and analyzing the laser scanning position height information Zc (x, y) and the depth camera sensor recognition position height Zl (x, y); S12. Marking the ground feature data set G with timestamp information and unique identification information of the load, thereby forming a collection feature set E of the ground of the robot's forward path for transporting the load at time t.

3. The method for controlling the running posture of a heavy-load robot according to claim 2, characterized in that: The laser scanning position height information Zc (x, y) is obtained by the laser scanning height information measured in real time by the laser scanner, and the depth camera sensor recognition position height Zl (x, y) is obtained by real-time analysis of the depth camera sensor; The height value Zg (x, y) is obtained by the following calculation formula: ； Wherein, max represents the maximum value function, and max(Zc(x, y), Zl(x, y)) represents the maximum height value of the laser scanning position height information Zc(x, y) and the depth camera sensor recognition position height Zl(x, y) at the same position (x, y); The slope change value Gd(x, y) is obtained by the following calculation formula: ； Wherein, arctan represents the inverse tangent function, and dx represents the horizontal distance between the height values ​​at the position (x, y) and the position (x+1, y); The obstacle identification ZAbs is obtained by comparing with a preset ground height threshold Zgthe of the robot's forward path; The comparison is made in the following ways: ； When the obstacle mark ZAbs ≥ the ground height threshold Zgthe of the robot's forward path, the height value of the robot's forward path ground at the position (x, y) is judged to be abnormal, and the obstacle mark ZAbs = 1; When the obstacle mark ZAbs is less than the ground height threshold Zgthe of the robot's forward path, it is determined that the height value of the ground of the robot's forward path at the position (x, y) is normal, and the obstacle mark ZAbs=0.

4. The method for controlling the running posture of a heavy-load robot according to claim 3, characterized in that: The S2 includes S21 and S22; S21. Obtain the load state of the robot by establishing a monitoring system integrated with the robot, including using an accelerometer and a force sensor to collaboratively obtain the load mass of each component of the robot, to obtain the center of gravity position G of the robot in the current operating state, and compare it with the preset target center of gravity position Gtarget of the robot to obtain the current center of gravity offset △G of the robot; S22. Integrate the center of gravity offset △G with the robot's running posture angle to output the robot's state evaluation vector B, where the running posture angle includes the pitch angle Pitch, the roll angle Roll and the heading angle Yaw, which are specifically obtained through the gyroscope sensor, the acceleration sensor and the magnetometer sensor.

5. The method for controlling the running posture of a heavy-load robot according to claim 4, characterized in that: The center of gravity position G is obtained by the following calculation formula: ； Where Xg, Yg and Zg represent the horizontal, vertical and vertical coordinates of the total center of gravity of the robot, n represents the total number of robot components, q(i) represents the mass of the i-th component, x(i), y(i) and z(i) represent the horizontal, vertical and vertical coordinates of the i-th component, respectively. The center of gravity offset ΔG is The calculation formula is obtained, wherein the target center of gravity position Gtarget includes the target horizontal axis Xtarget, the target vertical axis Ytarget and the target longitudinal axis Ztarget of the target center of gravity position.

6. The method for controlling the running posture of a heavy-load robot according to claim 5, characterized in that: The S3 includes S31; S31, by analyzing the influence of the robot's forward path ground on the robot's posture and load based on the collected feature set E and the state evaluation vector B, calculate and obtain the robot's adaptive posture adjustment vector △D, including the horizontal axis position offset △X, the vertical axis position offset △Y, the longitudinal axis position offset △Z, the pitch angle adjustment △Pitch, the roll angle adjustment △Roll and the heading angle adjustment △Yaw; The horizontal axis position offset △X is obtained by the following calculation formula: ； Wherein, k1 represents the first proportional coefficient, Zg(x, y) represents the height value at the position (x, y), and Gd(x, y) represents the slope change value at the position (x, y); The vertical axis position offset △Y is obtained by the following calculation formula: ； Wherein, k2 represents the second proportional coefficient, which is specifically used to adjust the influence of the height difference of the path on the vertical axis position offset △Y axis offset; The longitudinal axis position offset ΔZ is obtained by the following calculation formula: ； Wherein, k3 represents the third proportional coefficient, which is specifically used to control the influence of the obstacle on the longitudinal axis position offset △Z, and ZAbs represents the obstacle identifier; The pitch angle adjustment amount ΔPitch is obtained by the following calculation formula: ； Wherein, k4 and k7 represent the fourth proportional coefficient and the seventh proportional coefficient respectively, which are specifically used to adjust the influence of the offset on the pitch angle adjustment amount △Pitch; The roll angle adjustment amount ΔRoll is obtained by the following calculation formula: ； Wherein, k5 and k8 represent the fifth proportional coefficient and the eighth proportional coefficient respectively, which are specifically used to adjust the influence of the offset on the roll angle adjustment amount △Roll; The heading angle adjustment ΔYaw is obtained by the following calculation formula: ； Wherein, k6 and k9 represent the sixth proportional coefficient and the ninth proportional coefficient respectively, which are specifically used to adjust the influence of the offset on the heading angle adjustment amount △Yaw.

7. The method for controlling the running posture of a heavy-load robot according to claim 6, characterized in that: The S4 includes S41; S41, the robot adjusts the current forward posture of the robot based on the acquired posture adjustment vector △D, specifically by converting the horizontal axis position offset △X, the vertical axis position offset △Y, the longitudinal axis position offset △Z, the pitch angle adjustment △Pitch, the roll angle adjustment △Roll and the heading angle adjustment △Yaw in the posture adjustment vector △D into a control command, and the robot's internal motion control system acts on the mechanical parts to change the robot's forward posture, and the adjusted execution vector W is obtained, and the execution vector W is generated. The control command is used to dynamically adjust the robot's forward posture and mark the current robot's center of gravity position G; The execution vector W is obtained by the following conversion formula: ； Where f represents the transfer function preset by the control system.

8. The method for controlling the running posture of a heavy-load robot according to claim 7, characterized in that: The S5 includes S51; S51, by counting the fluctuation change information of the center of gravity position G of the robot within a fixed period, the center of gravity position change △Z is obtained, and compared with the preset robot operation center of gravity offset threshold Zthe, the current robot operation center of gravity offset evaluation result is obtained, and an iterative optimization mechanism is performed according to the operation center of gravity offset evaluation result; The center of gravity position change ΔZ is obtained by the following calculation formula: ； Where M represents the total number of sampling times within a fixed period, t(j+1, G) represents the centroid position at the j+1th sampling time t, and t(j, G) represents the centroid position at the jth sampling time t; The running center of gravity offset evaluation result is obtained by the following comparison method: When the center of gravity position change △Z ≥ the robot operation center of gravity offset threshold Zthe, the operation center of gravity offset evaluation result is obtained as an abnormal result, triggering the iterative optimization mechanism, including adjusting the acquisition of the posture adjustment vector △D, correcting the center of gravity position and the posture angle, specifically by adjusting the first proportional coefficient k1 to the ninth proportional coefficient k9; When the center of gravity position change △Z is less than the robot operation center of gravity offset threshold Zthe, the operation center of gravity offset evaluation result is obtained as a normal result.

9. A running posture control system for a heavy-load robot, applied to a running posture control method for a heavy-load robot as claimed in any one of claims 1 to 8, characterized in that: It includes forward path acquisition module, machine status acquisition module, posture analysis and adjustment module, adjustment generation and marking module and iterative optimization module; The forward path acquisition module acquires the real-time data of the ground of the current forward path of the robot in real time by integrating a laser scanner and a depth camera sensor on the robot, and records it as a collection feature set E after collection and analysis; The machine state acquisition module acquires the load distribution of the robot by establishing a monitoring system integrated with the robot, and calculates the center of gravity offset △G of the current robot according to the load state, and outputs the state evaluation vector B of the robot; The posture analysis and adjustment module analyzes the influence of the robot's forward path ground on the robot's posture and load by collecting feature set E and state evaluation vector B, and calculates and obtains the robot's adaptive posture adjustment vector △D; The adjustment generation and marking module adjusts the forward posture of the current robot based on the acquired posture adjustment vector △D through the robot, obtains the adjusted execution vector W, and generates a control command for the execution vector W to dynamically adjust the forward posture of the robot and mark the center of gravity position G of the current robot; The iterative optimization module obtains the current robot's operating center of gravity offset evaluation result by counting the robot's center of gravity position change △Z within a fixed period and comparing it with the preset robot operating center of gravity offset threshold Zthe, and performs iterative optimization based on the operating center of gravity offset evaluation result.

10. A storage medium for controlling the running posture of a heavy-load robot: the storage medium stores a computer program, and when the computer program is executed, it implements the method for controlling the running posture of a heavy-load robot as described in any one of claims 1 to 8.

Images