Intelligent power management and self-adaptive homeward voyage control method for cable-climbing robot
Through the integrated encoder, power margin monitoring module and deep learning algorithm, the problem of inaccurate power management of the cable robot is solved, and intelligent power management and adaptive return control are realized to ensure safe completion of tasks and improve energy efficiency.
Patent Information
- Application Number
- CN202411101752.7
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-08-12
- Publication Date
- 2024-11-22
AI Technical Summary
During high-altitude detection or maintenance tasks, the cable robot lacks accurate power consumption prediction models, resulting in the risk of premature return or power exhaustion, which affects the smooth completion of the task and the safe recycling of equipment.
Using integrated encoder, power margin monitoring module and deep learning algorithm, intelligent power management and adaptive return by collecting robot power residual amount, fuselage weight, cable rod inclination and motor power data, combined with deep cyclic neural network model Control, optimize operational efficiency and ensure safe return.
Accurately predict the power supply margin, avoid unnecessary early return, ensure timely and safe return when low power is low, reduce the risks of high-altitude operations, maximize the utilization of electricity, improve operational energy efficiency, realize the automation of power management and return decisions, and reduce the burden of manual monitoring .
Smart Images

Figure CN119002274A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the field of power management of a cable-climbing robot, and in particular to an intelligent power management and adaptive return control method for a cable-climbing robot. Background Art
[0002] When performing high-altitude inspection or maintenance tasks, the cable-climbing robot is limited by the limited battery capacity, and its endurance is directly related to the successful completion of the task and the safe recovery of the equipment. Traditional methods lack an accurate power consumption prediction model, which often leads to premature return and failure to fully utilize the battery energy, or the risk of running out of power at high altitude due to insufficient estimation, which brings great challenges to equipment recovery. Therefore, it is necessary to design an intelligent power management and adaptive return control method for cable-climbing robots. Summary of the invention
[0003] The purpose of the present invention is to provide an intelligent power management and adaptive return control method for a rope-climbing robot, so as to solve the technical problems of the existing rope-climbing robot's power management being inaccurate, returning to work prematurely or running out of energy.
[0004] By integrating encoders and power reserve monitoring modules and combining them with deep learning algorithms, intelligent management of the robot's power status and adaptive return control strategies are achieved to optimize operating efficiency and ensure operational safety.
[0005] In order to achieve the above object, the technical solution adopted by the present invention is as follows:
[0006] Intelligent power management and adaptive return control method for a rope-climbing robot, the method comprising the following steps:
[0007] Step 1: Collect the remaining amount of power of the rope-climbing robot;
[0008] Step 2: Obtain data on fuselage weight, cable rod inclination, and motor power;
[0009] Step 3: Input the acquired data and remaining power into a pre-trained deep learning model to obtain the possible driving distance;
[0010] Step 4: Then process the actual distance traveled and the distance that can be traveled based on the initial collection calculation, and then divide it by 2 to get the current actual distance that can be traveled.
[0011] Furthermore, in step 1, the current power supply of the climbing robot is collected in real time. The collection is performed before crawling to determine the initial remaining power supply. No crawling distance budget is performed at the initial power level. Then, after collecting the set fixed crawling distance, the actual crawling distance is collected and the power level is collected again. Then, the distance that can be crawled with the remaining power is budgeted based on the electric energy consumed in the actual crawling distance.
[0012] Furthermore, in step 2, the body weight is the initial setting value, which is manually set according to the actual weight of the climbing robot. The inclination of the cable rod is based on the actual crawling distance and the actual ascent height of the climbing robot at two time points, and then the third side of the triangle, that is, the horizontal distance, is obtained according to the Pythagorean theorem. Then the inclination of the cable rod can be calculated, and the cable rod is divided into several sections. Then the inclination obtained is also several inclinations. The inclination of the future distance is predicted according to the characteristics of the locking rod, so that the power energy consumption is more accurate.
[0013] Furthermore, in step 3, the fuselage weight, the inclination of the cable rod, the motor power and the remaining amount of the power supply are used as input parameters of the training model, and then the actual crawling distance that can be achieved with the power is calculated based on the model.
[0014] Furthermore, the cable-climbing robot is also provided with an unexpected recovery mechanism. The crawling support rod of the cable-climbing robot is arranged as a retractable structure. The pressure between the crawling wheel and the locking rod is determined according to the telescopic length, and then the friction between the cable rod and the crawling wheel is collected. At the same time, a reserved power supply for controlling the retractable structure is arranged. When the power supply fails to return home due to an accident, the retractable structure is controlled to shorten and relax, so that the crawling wheel slides and crawls under the gravity of the cable-climbing robot, and returns home automatically by relying on the gravity of the cable-climbing robot. At the same time, the energy of the reverse rotation of the crawling wheel is recovered to the power supply to control the retractable structure, so that the cable-climbing robot returns home automatically when there is an accident with the power supply.
[0015] Furthermore, when calculating the actual distance that can be advanced, timed data collection is performed. The collection condition is that after the crawling wheel crawls a fixed distance, the actual distance that can be advanced is calculated. Since the inclination of the cable rod is variable, the actual crawling distance is different at different inclinations. Therefore, the inclination of the most recent crawling is used to predict the inclination of the next stage, and the predicted inclination of the next stage is used as a parameter of the deep learning model to make the accuracy more precise.
[0016] Furthermore, a deep recurrent neural network is used to train a deep learning model. The model input includes but is not limited to the crawling distance, power reserve, robot weight, motor power consumption, cable rod inclination angle, etc., and the output is the estimated crawling distance.
[0017] Based on the model prediction results and the preset safety threshold (such as the minimum return power), the return trigger point is calculated in real time. If the remaining power is not enough to support the robot to return to the starting point safely, the return program is started immediately.
[0018] Assuming that the current crawling distance is s1, and the distance that can be continued to travel output by the model is L, the actual distance that the climbing robot can continue to travel is S = (L-s1) / 2. In order to ensure that the climbing robot can return safely, it is necessary to set the minimum value of S. The allowable minimum value is the return trigger point. The trigger point is 1, that is, when S is 1m, theoretically it can continue to move forward 1m, but for safety reasons, it should return immediately. On the premise of ensuring safe return, the algorithm further optimizes the crawling path to maximize the use of remaining power for effective detection or operation.
[0019] The present invention has the following beneficial effects due to the adoption of the above technical solution:
[0020] The present invention accurately predicts the remaining power supply, avoids unnecessary early return, and ensures timely and safe return when the power is low, reduces the risk of high-altitude operations, manages the power supply in a refined manner, maximizes the use of each unit of electrical energy, improves the energy efficiency of the overall operation, realizes the automation of power supply management and return decision-making, reduces the burden of manual monitoring, and improves the level of operation automation. BRIEF DESCRIPTION OF THE DRAWINGS
[0021] Figure 1 It is a flow chart of the method of the present invention. DETAILED DESCRIPTION
[0022] In order to make the purpose, technical solution and advantages of the present invention more clearly understood, the present invention is further described in detail below with reference to the accompanying drawings and preferred embodiments. However, it should be noted that many details listed in the specification are only for the purpose of enabling the reader to have a thorough understanding of one or more aspects of the present invention, and these aspects of the present invention can be implemented even without these specific details.
[0023] like Figure 1 As shown, the intelligent power management and adaptive return control method of the climbing robot comprises the following steps:
[0024] Step 1: Collect the remaining power of the current crawling robot. The current power of the crawling robot is collected in real time. The collection is performed before crawling to determine the initial remaining power. The crawling distance is not budgeted at the initial power level. After the set crawling distance is collected, the actual crawling distance is collected and the power level is collected. Then, the remaining power is budgeted according to the power consumed by the actual crawling distance.
[0025] Step 2: Obtain the data of the fuselage weight, cable rod inclination, and motor power. The fuselage weight is the initial setting value, which is manually set according to the actual weight of the cable climbing robot. The cable rod inclination is based on the actual crawling distance and the actual height of the cable climbing robot at two time points, and then the third side of the triangle, that is, the horizontal distance, is obtained according to the Pythagorean theorem. Then the cable rod inclination can be calculated, and the cable rod is divided into several sections. Then the inclination obtained is also several inclinations. The inclination of the future distance is predicted according to the characteristics of the locking rod, making the power energy consumption more accurate.
[0026] Step 3: Input the acquired data and the remaining power into the pre-trained deep learning model to obtain the distance that can be traveled. The fuselage weight, cable rod inclination, motor power and remaining power are used as input parameters of the training model, and then the model is used to calculate the actual crawling distance that can be traveled with the power.
[0027] The deep learning model is trained using a deep recurrent neural network. The model input includes but is not limited to the crawling distance, power reserve, robot weight, motor power consumption, cable rod inclination angle, etc. The output is the estimated crawling distance.
[0028] Based on the model prediction results and the preset safety threshold (such as the minimum return power), the return trigger point is calculated in real time. If the remaining power is not enough to support the robot to return to the starting point safely, the return program is started immediately.
[0029] Assuming that the current crawling distance is s1, and the distance that can be continued to travel output by the model is L, the actual distance that the climbing robot can continue to travel is S = (L-s1) / 2. In order to ensure that the climbing robot can return safely, it is necessary to set the minimum value of S. The allowable minimum value is the return trigger point. The trigger point is 1, that is, when S is 1m, theoretically it can continue to move forward 1m, but for safety reasons, it should return immediately. On the premise of ensuring safe return, the algorithm further optimizes the crawling path to maximize the use of remaining power for effective detection or operation.
[0030] Step 4: Then process the actual distance traveled and the distance that can be traveled calculated by initial collection, and then divide it by 2 to get the current actual distance that can be traveled. When calculating the actual distance that can be traveled, timed collection is performed. The collection condition is that after the crawling wheel crawls a fixed distance, the actual distance that can be traveled is calculated once. Since the inclination of the cable rod is variable, the actual distance that can be crawled is different for different inclinations. Therefore, the inclination of the most recent crawling is used to predict the inclination of the next stage, and the predicted inclination of the next stage is used as the parameter of the deep learning model to make the accuracy more precise.
[0031] The cable-climbing robot is also provided with an accidental recovery mechanism. The crawling support rod of the cable-climbing robot is set as a retractable structure. The pressure between the crawling wheel and the locking rod is determined according to the telescopic length, and then the friction between the cable rod and the crawling wheel is collected. At the same time, a reserved power supply is set to control the retractable structure. When the power supply is unexpectedly low and cannot return, the retractable structure is controlled to shorten and relax, so that the crawling wheel slides and crawls under the gravity of the cable-climbing robot, and returns automatically by relying on the gravity of the cable-climbing robot. At the same time, the energy from the reverse rotation of the crawling wheel is recovered to the power supply to control the retractable structure, so that the cable-climbing robot automatically returns when the power supply is unexpected.
[0032] The cable-climbing robot is equipped with an encoder, a power reserve monitoring module, and a prediction model based on deep learning. The system can monitor the crawling distance and power status in real time, and dynamically calculate the distance that can be continued to work in combination with other key parameters, and automatically trigger the safe return mechanism. Encoder integration: used to monitor the crawling distance of the cable-climbing robot in real time. Power reserve monitoring module: monitors and reports the current battery power status in real time. Other sensors: including but not limited to body weight sensors, motor power monitors, and cable rod inclination sensors, providing more dimensional input data for the algorithm model.
[0033] Improve operational efficiency and safety: Accurately predict the remaining power supply to avoid unnecessary early return, while ensuring timely and safe return when the power is low, reducing the risk of high-altitude operations. Optimize energy utilization: Finely manage power supply, maximize the use of each unit of electricity, and improve the energy efficiency of the overall operation. Automated decision-making: Automated power management and return decision-making, reduce the burden of manual monitoring, and improve the level of operational automation. By integrating multiple sensor data and deep learning algorithms, the crawling distance that can be supported by the remaining power can be predicted in real time to achieve the optimal operation path planning and safe return decision-making.
[0034] The above is only a preferred embodiment of the present invention. It should be pointed out that for ordinary technicians in this technical field, several improvements and modifications can be made without departing from the principle of the present invention. These improvements and modifications should also be regarded as the scope of protection of the present invention.
Claims
1. Intelligent power management and adaptive return control method for a rope-climbing robot, characterized by: The method comprises the following steps: Step 1: Collect the remaining amount of power of the rope-climbing robot; Step 2: Obtain data on fuselage weight, cable rod inclination, and motor power; Step 3: Input the acquired data and remaining power into a pre-trained deep learning model to obtain the possible driving distance; Step 4: Then process the actual distance traveled and the distance that can be traveled based on the initial collection calculation, and then divide it by 2 to get the current actual distance that can be traveled.
2. The intelligent power management and adaptive return control method of the rope-climbing robot according to claim 1 is characterized by: In step 1, the current power supply of the climbing robot is collected in real time. The collection is performed before crawling to determine the initial remaining power supply. No crawling distance budget is performed at the initial power level. Then, after collecting the set fixed crawling distance, the actual crawling distance is collected and the power level is collected again. Then, the distance that can be crawled with the remaining power is budgeted based on the electric energy consumed in the actual crawling distance.
3. The intelligent power management and adaptive return control method of the rope-climbing robot according to claim 1 is characterized by: In step 2, the body weight is the initial setting value, which is manually set according to the actual weight of the climbing robot. The inclination of the cable rod is based on the actual crawling distance and the actual ascent height of the climbing robot at two time points, and then the third side of the triangle is obtained according to the Pythagorean theorem, that is, the horizontal distance. Then the inclination of the cable rod can be calculated, and the cable rod is divided into several sections. Then the inclination obtained is also several inclinations. The inclination of the future distance is predicted according to the characteristics of the locking rod, so that the power energy consumption is more accurate.
4. The intelligent power management and adaptive return control method of the rope-climbing robot according to claim 1 is characterized in that: In step 3, the fuselage weight, cable rod inclination, motor power and remaining power supply are used as input parameters of the training model, and then the actual crawling distance that can be achieved with the power is calculated based on the model.
5. The intelligent power management and adaptive return control method of the rope-climbing robot according to claim 1 is characterized in that: The cable-climbing robot is also provided with an accidental recovery mechanism. The crawling support rod of the cable-climbing robot is set as a retractable structure. The pressure between the crawling wheel and the locking rod is determined according to the telescopic length, and then the friction between the cable rod and the crawling wheel is collected. At the same time, a reserved power supply is set to control the retractable structure. When the power supply is unexpectedly low and cannot return, the retractable structure is controlled to shorten and relax, so that the crawling wheel slides and crawls under the gravity of the cable-climbing robot, and returns automatically by relying on the gravity of the cable-climbing robot. At the same time, the energy from the reverse rotation of the crawling wheel is recovered to the power supply to control the retractable structure, so that the cable-climbing robot automatically returns when the power supply is unexpected.
6. The intelligent power management and adaptive return control method of the rope-climbing robot according to claim 1 is characterized by: When calculating the actual distance that can be advanced, timed data collection is performed. The collection condition is that after the crawling wheel crawls a fixed distance, the actual distance that can be advanced is calculated. Since the inclination of the cable rod is variable, the actual crawling distance is different for different inclinations. Therefore, the inclination of the most recent crawling is used to predict the inclination of the next stage. The predicted inclination of the next stage is used as the parameter of the deep learning model to make the accuracy more precise.
7. The intelligent power management and adaptive return control method of the rope-climbing robot according to claim 4 is characterized by: The deep learning model is trained using a deep recurrent neural network. The model input includes but is not limited to the crawling distance, power reserve, robot weight, motor power consumption, cable rod inclination angle, etc. The output is the estimated crawling distance. Based on the model prediction results and the preset safety threshold or minimum return power, it calculates in real time whether the return trigger point has been reached. If the remaining power is not enough to support the robot to return to the starting point safely, the return program is started immediately. Assuming that the current crawling distance is s1, and the distance that can be continued to travel output by the model is L, the actual distance that the climbing robot can continue to travel is S = (L-s1) / 2. In order to ensure that the climbing robot can return safely, it is necessary to set the minimum value of S. The allowable minimum value is the return trigger point. The trigger point is 1, that is, when S is 1m, theoretically it can continue to move forward 1m, but for safety reasons, it should return immediately. On the premise of ensuring safe return, the remaining power can be maximized for effective detection or operation.
