Hierarchical control AGV robot obstacle avoidance method and system

Abstract

The invention provides a hierarchical control AGV robot obstacle avoidance method and system, and relates to the field of general control or regulation systems.The method comprises the steps that AGV robot operation parameters are collected and input into a risk assessment model, and risk grade values are output; determining an obstacle avoidance execution mechanism triggered first according to a preset interval in which the risk level value is located, wherein the obstacle avoidance execution mechanism comprises a hardware control layer, a local calculation layer and a remote collaboration layer; and when the local calculation layer is triggered, an obstacle avoidance area of the AGV robot in a preset range is divided into 16 fan-shaped spaces, the target fan-shaped space with the highest traffic index is determined as an avoidance direction, and a corresponding target obstacle avoidance scheme is selected according to channel width parameters of the target fan-shaped space. By implementing the method, the obstacle avoidance response time can be shortened, the processing delay caused by the fact that all obstacle avoidance decisions in a traditional scheme need the participation of the main control system is avoided, and the collision risk caused by waiting for the processing of the main control system is reduced.

Description

Technical Field

[0001] The present application relates to the general field of control or regulation systems, and in particular to a hierarchically controlled AGV robot obstacle avoidance method and system. Background Art

[0002] With the rapid development of industrial automation, AGVs (Automated Guided Vehicles) (AGVs) are increasingly being used in factories, warehouses, and other scenarios. In complex industrial environments, AGVs must respond to various emergencies and ensure the safe and efficient completion of handling tasks. Therefore, obstacle avoidance is a key technology for AGVs to ensure their normal operation.

[0003] In related technologies, obstacle avoidance control for AGVs is primarily implemented through centralized control. AGVs collect environmental information using visual sensors (such as cameras) and lidar, transmitting this information to a central control system for analysis. Upon receiving an abnormality signal, the control system first verifies the authenticity of the abnormality, then generates obstacle avoidance instructions based on pre-set scheduling rules. Finally, these instructions are issued to the AGVs to execute the corresponding avoidance maneuvers. This obstacle avoidance approach, relying on the unified scheduling of the control system, facilitates the collaborative operation of multiple AGVs.

[0004] However, in practice, the acquisition, transmission, confirmation, and command execution of abnormal signals all require processing time, and network transmission may be delayed. This often results in a significant delay between the time an AGV detects an obstacle and the time it executes an avoidance maneuver. This processing delay can make it difficult for the AGV to make timely avoidance maneuvers, especially when operating at high speeds or encountering unexpected obstacles, increasing the risk of a collision. Summary of the Invention

[0005] The present application provides a hierarchically controlled AGV robot obstacle avoidance method and system, which is used to address the problem of how to enable the AGV robot to quickly make avoidance actions after detecting an obstacle, thereby reducing the risk of collision caused by waiting for the main control system to process.

[0006] In a first aspect, the present application provides a hierarchically controlled AGV robot obstacle avoidance method, which is applied to an AGV robot control system. The AGV robot control system includes three obstacle avoidance actuators: a hardware control layer, a local computing layer, and a remote collaboration layer. The method includes: Collecting AGV robot operating parameters, including load value, ground friction coefficient, obstacle distance value, historical accident data and mission time parameters; Inputting the operating parameters into a risk assessment model and outputting a risk level value; The obstacle avoidance actuator to be triggered first is determined based on the preset interval in which the risk level value is located. The preset interval includes a first preset interval corresponding to the hardware control layer, a second preset interval of the local computing layer, and a third preset interval of the remote collaboration layer. The hardware control layer is used to perform emergency braking on the AGV robot, and the remote collaboration layer is connected to the main control system. When the local computing layer is triggered, the obstacle avoidance area of ​​the AGV robot within the preset range is divided into 16 sector spaces, and the traffic index of each sector space is calculated according to the traffic index calculation formula; The target sector space with the highest traffic index is determined as the avoidance direction, and a corresponding target obstacle avoidance solution is selected according to the channel width parameter of the target sector space.

[0007] Through the above-mentioned embodiments, the AGV robot control system achieves hierarchical rapid response for the AGV robot by establishing a three-level obstacle avoidance mechanism consisting of a hardware control layer, a local computing layer, and a remote collaboration layer. The system can automatically trigger the most appropriate obstacle avoidance level based on the risk level, without having to wait for instructions from the main control system each time. When the risk is low, the local computing layer can independently complete obstacle avoidance decisions and quickly select the optimal obstacle avoidance path through 16-sector space analysis and traffic index evaluation. This hierarchical control strategy reduces obstacle avoidance response time and avoids the processing delay caused by the traditional solution where all obstacle avoidance decisions require the participation of the main control system, effectively reducing the risk of collision caused by waiting for the main control system to process.

[0008] In some embodiments, the step of inputting the operating parameters into a risk assessment model and outputting a risk level value specifically includes: The hardware control layer performs a threshold judgment based on the load value, the ground friction coefficient, and the obstacle distance value to obtain a first risk index; The local computing layer inputs the first risk index, the historical accident data, and the task time parameter into a fuzzy evaluator, and calculates a second risk index based on a preset fuzzy rule; The remote collaboration layer inputs the first risk index and the second risk index into a pre-trained deep learning model to obtain a third risk index; The risk level value is determined according to a weighted average of the first risk index, the second risk index, and the third risk index.

[0009] Through the above-described embodiment, the hardware control layer can quickly determine whether emergency obstacle avoidance is necessary based on basic parameters, without waiting for confirmation from the master control system. The local computing layer uses a fuzzy evaluator to quickly assess risk. The remote collaboration layer provides more accurate risk assessments when conditions permit. This layered risk assessment mechanism ensures that the corresponding obstacle avoidance process can be initiated as quickly as possible at different risk levels, effectively overcoming the response delays caused by traditional solutions that rely too heavily on the master control system for risk assessment.

[0010] In some embodiments, the step of dividing the obstacle avoidance area of ​​the AGV robot within a preset range into 16 sector-shaped spaces specifically includes: Calculating the motion characteristic coefficient of the AGV robot based on the speed parameters of the AGV robot, wherein the speed parameters include current speed, steering angle, and acceleration; Dividing the obstacle avoidance area within the preset range into a forward area, a side area, and a rearward area according to the motion characteristic coefficient, wherein the forward area, the side area, and the rearward area correspond to 8 sector spaces, 6 sector spaces, and 2 sector spaces, respectively; The spatial weight coefficient is calculated according to the sector-shaped spatial distribution, and the spatial weight coefficient is used as a weighting parameter in the traffic index calculation formula.

[0011] Through the above-described embodiment, the AGV robot control system divides the obstacle avoidance area into eight sectors in the forward direction, six sectors in the lateral direction, and two sectors in the rearward direction, enabling the AGV robot to quickly assess obstacle avoidance possibilities in all directions. This non-uniform spatial division scheme, based on motion characteristics, emphasizes the importance of the forward and lateral sectors, enabling the AGV robot to complete spatial analysis and select the optimal avoidance direction in the shortest possible time. This solution eliminates the need for the main control system to perform complex path planning, significantly shortening obstacle avoidance decision-making time and improving the AGV robot's ability to quickly respond to unexpected obstacles.

[0012] In some embodiments, the step of calculating the traffic index of each sector space according to the traffic index calculation formula specifically includes: Collecting real-time status parameters of each sector space and calculating an instantaneous traffic index based on the real-time status parameters, wherein the real-time status parameters include obstacle density, moving speed, and relative distance; Based on the changing trend of the real-time state parameters in the historical sampling period, a Kalman filter algorithm is used to predict the state evolution of each sector space within a preset time window to obtain a short-term traffic index; Establishing a long-term evaluation model based on the historical traffic records of each sector space and calculating the long-term traffic index corresponding to each sector space; The weight coefficients corresponding to the instantaneous traffic index, the short-term traffic index and the long-term traffic index are adjusted respectively according to the complexity of the current scene, and the traffic index is obtained through weighted combination.

[0013] Through the above-described embodiments, the AGV robot control system calculates the instantaneous pass index in real time, enabling the AGV robot to respond immediately to changes in its surroundings. Short-term predictions using the Kalman filter algorithm can predict obstacle movement trends. Long-term evaluations combined with historical data provide a more reliable basis for decision-making. This multi-dimensional local evaluation mechanism enables the AGV robot to independently make most obstacle avoidance decisions, significantly reducing its reliance on the master control system and avoiding the safety risks associated with communication delays.

[0014] In some embodiments, the step of selecting a corresponding target obstacle avoidance solution according to the channel width parameter of the target sector space specifically includes: Calculating an obstacle avoidance level based on the channel width parameter of the target sector space, and determining an initial obstacle avoidance plan based on the obstacle avoidance level, wherein the obstacle avoidance levels are deceleration obstacle avoidance, lane change obstacle avoidance, and emergency obstacle avoidance from low to high; After the initial obstacle avoidance plan is started, if the obstacle avoidance effect parameters of the AGV robot are detected to be lower than the preset target threshold, the plan upgrade judgment is triggered. The obstacle avoidance effect parameters include relative distance change rate, heading deviation and obstacle avoidance progress; In the scheme upgrade judgment, when it is detected that the obstacle avoidance margin index is greater than a preset upgrade threshold, the initial obstacle avoidance scheme is upgraded to a target obstacle avoidance scheme corresponding to a higher obstacle avoidance level. The obstacle avoidance margin index is determined based on the remaining decision time and maneuvering space.

[0015] Through the above-described embodiment, the AGV robot control system directly selects an initial obstacle avoidance level based on aisle width and automatically upgrades the plan based on the effectiveness of the avoidance strategy, eliminating the need to consult the master control system each time. Introducing the obstacle avoidance margin index as the upgrade criteria ensures that the obstacle avoidance level is immediately increased as the risk level increases. This adaptive local decision-making mechanism significantly improves the AGV robot's response speed to emergencies and effectively prevents missed obstacle avoidance opportunities due to waiting for a decision from the master control system.

[0016] In some embodiments, after the step of determining the obstacle avoidance actuator to be triggered first based on the preset interval of the risk level value, the method further includes: When the hardware control layer is triggered, emergency braking is immediately performed; After the emergency braking, the vehicle resumes driving according to the control instructions issued by the main control system.

[0017] Through the above-described embodiment, when an emergency situation is detected, the AGV robot can immediately perform emergency braking without waiting for confirmation from the master control system, allowing it to take evasive action in the shortest possible time. This mechanism eliminates the response delay caused by traditional solutions requiring authorization from the master control system, providing a final safeguard for the AGV robot. Furthermore, by requiring authorization from the master control system before resuming travel, safe recovery after an emergency obstacle avoidance is ensured.

[0018] In some embodiments, after the step of determining the obstacle avoidance actuator to be triggered first based on the preset interval of the risk level value, the method further includes: When the remote collaboration layer is triggered, the AGV robot is controlled to wait for the control instructions issued by the main control system during normal driving; The AGV robot is controlled to avoid obstacles according to the control instructions.

[0019] Through the above-described embodiments, the AGV robot control system has designed a special obstacle avoidance mechanism for the remote collaboration layer, enabling the AGV robot to maintain normal operation while awaiting instructions from the master control system. This mechanism avoids the chain reaction that could otherwise be triggered by simply stopping and waiting, minimizing the impact of waiting for the master control system while ensuring safety. Furthermore, by reserving a remote collaboration interface, the system can leverage the master control system's global perspective to make more optimized obstacle avoidance decisions when necessary, achieving an organic combination of rapid response and global optimization.

[0020] In a second aspect, the present application provides an AGV robot control system, the AGV robot control system comprising: one or more processors and a memory; The memory is coupled to the one or more processors, and the memory is used to store computer program code, and the computer program code includes computer instructions. The one or more processors call the computer instructions so that the AGV robot control system can implement a hierarchically controlled AGV robot obstacle avoidance method provided in the above embodiment, which will not be repeated here.

[0021] On the third aspect, the present application provides a computer-readable storage medium, including instructions. When the instructions are run on an AGV robot control system, the AGV robot control system can implement a hierarchically controlled AGV robot obstacle avoidance method provided in the above embodiment, which will not be repeated here.

[0022] Fourthly, the present application provides a computer program product. When the computer program product runs on an AGV robot control system, the AGV robot control system can implement a hierarchically controlled AGV robot obstacle avoidance method provided in the above embodiment, which will not be repeated here.

[0023] One or more technical solutions provided in the embodiments of this application have at least the following technical effects or advantages: 1. By establishing a hierarchical control architecture consisting of hardware control, local computing, and remote collaboration layers, the AGV robot control system can automatically trigger the most appropriate obstacle avoidance level based on risk level. The hardware control layer can directly execute emergency braking, the local computing layer can independently complete general obstacle avoidance decisions, and the remote collaboration layer handles situations requiring global coordination. This hierarchical mechanism eliminates the need for AGV robots to constantly wait for instructions from the master control system, allowing them to autonomously select the fastest response path based on risk level, fundamentally eliminating the processing delays inherent in traditional centralized control solutions.

[0024] 2. The area surrounding the AGV robot is divided into 16 sectors, with a non-uniform distribution based on its motion characteristics: eight sectors for the forward direction, six sectors for the side, and two sectors for the rear. By combining a multi-dimensional traffic index calculation method based on real-time status, short-term prediction, and long-term evaluation, the AGV robot can quickly and locally analyze its surroundings and select the optimal obstacle avoidance path. This spatial division and evaluation system fully considers the AGV's motion characteristics, enabling it to independently make effective obstacle avoidance decisions in the shortest possible time.

[0025] 3. A hierarchical obstacle avoidance system has been established, encompassing deceleration, lane change, and emergency avoidance. The AGV control system quickly selects the initial obstacle avoidance level based on aisle width and automatically upgrades the plan based on avoidance effectiveness and the obstacle avoidance margin index. This adaptive mechanism enables the AGV to flexibly adjust its obstacle avoidance strategy based on actual conditions, ensuring both rapid response and effective avoidance. This effectively addresses the problem of delayed obstacle avoidance in traditional solutions, often caused by single solutions or complex decision-making processes. BRIEF DESCRIPTION OF THE DRAWINGS

[0026] Figure 1 This is a flow chart of a hierarchically controlled AGV robot obstacle avoidance method according to an embodiment of the present application; Figure 2 This is another flow chart of a hierarchically controlled AGV robot obstacle avoidance method according to an embodiment of the present application; Figure 3 This is a schematic diagram of the physical device structure of the AGV robot control system in an embodiment of the present application. DETAILED DESCRIPTION

[0027] The terms used in the following examples of the present application are only for the purpose of describing specific embodiments and are not intended to limit the present application. As used in the specification and appended claims of this application, the singular expressions "a," "an," "said," "above," "the," and "this" are intended to include plural expressions as well, unless the context clearly indicates otherwise. It should also be understood that the term "and / or" used in this application refers to any or all possible combinations comprising one or more of the listed items.

[0028] In the following, the terms "first" and "second" are used for descriptive purposes only and should not be understood to imply or suggest relative importance or implicitly indicate the number of the technical features indicated. Therefore, the features defined as "first" and "second" may explicitly or implicitly include one or more of the features. In the description of the embodiments of this application, unless otherwise specified, "plurality" means two or more.

[0029] For ease of understanding, the following describes the process of the method provided by this implementation. Figure 1 , which is a flow chart of a hierarchically controlled AGV robot obstacle avoidance method in an embodiment of the present application.

[0030] S101. Collect AGV robot operating parameters.

[0031] The AGV robot control system collects operating parameters in real time through a multi-sensor network, including load values, ground friction coefficient, obstacle distance values, historical accident data, and task time parameters. Load values ​​are collected through weight sensors on the AGV robot platform, which can monitor the current load status in real time. The ground friction coefficient is measured by special ground detection sensors that can adapt to different road conditions. Obstacle distance values ​​are obtained by collecting multi-source data using devices such as lidar, ultrasonic sensors, and visual sensors installed on the AGV robot, thereby building a comprehensive obstacle detection network. Historical accident data is stored in a local database and includes information on dimensions such as accident type, cause, and impact. Task time parameters include key indicators such as task urgency and remaining execution time, with task urgency being inversely correlated with remaining execution time.

[0032] S102: Input the operating parameters into the risk assessment model and output the risk level value.

[0033] Specifically, the hardware control layer performs threshold judgment based on the load value, ground friction coefficient and obstacle distance value to obtain the first risk index; the local computing layer inputs the first risk index, historical accident data and task time parameters into the fuzzy evaluator, and calculates the second risk index based on the preset fuzzy rules; the remote collaboration layer inputs the first risk index and the second risk index into the pre-trained deep learning model to obtain the third risk index; finally, the risk level value is determined based on the weighted average of the first risk index, the second risk index and the third risk index.

[0034] Optionally, the hardware control layer uses a threshold judgment algorithm to calculate the first risk index. The algorithm sets three safety thresholds: load safety threshold (such as 80% of the maximum load), friction coefficient safety threshold (such as 0.4), and obstacle distance safety threshold (such as 1.5 times the emergency braking distance). When any parameter exceeds the threshold, the hardware layer risk assessment is triggered. Among them, the first risk index The calculation formula is: , where W is the load percentage, μ is the friction coefficient, D is the obstacle distance, α1=0.3, α2=0.4, α3=0.3.

[0035] The local computation layer constructs a fuzzy evaluator, which contains three input variables ( , historical accident rate, mission urgency) and an output variable (second risk index ). Fuzzy rules use Mamdani-type reasoning, for example: if High and historical accident rate is high and mission urgency is high, then The membership function of the input variables adopts triangular distribution, and the output variables are defuzzified using the centroid method.

[0036] The remote collaborative layer deploys a pre-trained LSTM neural network model with the input and , the output is the third risk index The LSTM neural network model is trained on historical data and consists of three hidden layers with 128 neurons in each layer. The activation function is ReLU and the loss function is mean square error. The final risk level value is calculated by weighted average: R=0.5 +0.3 + 0.2 .

[0037] Taking an industrial scenario as an example, hardware layer computing = 0.7 (load 75%, friction coefficient 0.5, obstacle distance 1.2 meters), local layer calculated based on historical accident rate (0.5 times / hour) and mission urgency (high) =0.8, remote layer model prediction =0.85. The final risk level value R=0.7*0.5+0.8*0.3+0.85*0.2=0.75, which is a medium risk level.

[0038] It should be noted that the fuzzy logic controller contains 5 Fuzzy subsets (very low, low, medium, high, very high), 3 historical accident rate subsets (low, medium, high), and 3 task urgency subsets (low, medium, high), a total of 5×3×3=45 fuzzy rules are generated.

[0039] S103: Determine the obstacle avoidance actuator to be triggered first according to the preset interval of the risk level value.

[0040] Optionally, the preset intervals are divided into three levels: the first interval (R ≥ 0.9) triggers the hardware control layer, the second interval (0.7 ≤ R < 0.9) triggers the local computing layer, and the third interval (R < 0.7) triggers the remote collaboration layer. Optionally, the hardware control layer directly controls the AGV robot's motor braking system via PWM signals, with a response time of less than 50ms. The local computing layer activates the obstacle avoidance decision module, which includes submodules such as sector space division and traffic index calculation. The remote collaboration layer connects to the main control system via a 5G communication module and uses the MQTT protocol for data exchange.

[0041] When the risk level value R=0.85, the local computing layer is triggered. The AGV robot control system immediately stops transmitting data to the remote collaboration layer and starts the local obstacle avoidance decision-making process.

[0042] S104. When the local computing layer is triggered, the obstacle avoidance area of ​​the AGV robot within the preset range is divided into 16 sector spaces, and the traffic index of each sector space is calculated according to the traffic index calculation formula.

[0043] Specifically, when the local computing layer is triggered, the AGV robot control system calculates the motion characteristic coefficient of the AGV robot based on speed parameters such as the current speed, steering angle and acceleration of the AGV robot; then, according to the motion characteristic coefficient, the obstacle avoidance area within the preset range is divided into a forward area (including 8 sector-shaped spaces), a lateral area (including 6 sector-shaped spaces) and a backward area (including 2 sector-shaped spaces); finally, the space weight coefficient is calculated based on the sector-shaped space distribution, and the space weight coefficient is used as the weighted parameter in the pass index calculation formula.

[0044] Among them, the motion characteristic coefficient K is calculated using the formula: , where v is the current speed, g is the acceleration due to gravity, μ is the friction coefficient, and L is the wheelbase. The forward area is divided into eight sectors, each 45°; the lateral area is divided into six sectors, each 60°; and the rearward area is divided into two sectors, each 45°. Spatial weight coefficients are set based on the importance of the area: the forward sector has a weight of 0.4, the lateral sector has a weight of 0.3, and the rearward sector has a weight of 0.3.

[0045] Next, the AGV robot control system collects real-time state parameters such as obstacle density, moving speed and relative distance of each sector space, and calculates the instantaneous pass index I based on the real-time state parameters; then, based on the changing trend of the real-time state parameters within the historical sampling period, the Kalman filter algorithm is used to predict the state evolution of each sector space within the preset time window to obtain the short-term pass index S; then, a long-term evaluation model is established based on the historical pass records of each sector space, and the long-term pass index L corresponding to each sector space is calculated; finally, according to the complexity of the current scene, the weight coefficients corresponding to the instantaneous pass index, short-term pass index and long-term pass index are adjusted respectively, and the pass index P is obtained through weighted combination.

[0046] Specifically, the traffic index calculation adopts a multi-dimensional evaluation model. Among them, the calculation formula of the instant traffic index I is: , where ρ is the obstacle density, is the relative speed, and d is the relative distance.

[0047] The final circulation index calculation formula is: ,in, 、 and They represent the weight coefficients of I, S and L respectively, and the weight coefficients can be dynamically adjusted according to the complexity of the scene, which is not limited here.

[0048] Taking a logistics center as an example, the AGV robot's current speed is 2m / s, the wheelbase is 0.8m, and the friction coefficient is 0.6, so K is calculated to be 0.85. The obstacle avoidance area is divided into 8 sectors in the front direction (0-360°, one every 45°), 6 sectors on the side (3 sectors on the left or right, one every 60°), and 2 sectors in the rear direction. In real time, the obstacle density in the third sector in the front direction is detected to be 0.3, the relative speed is 0.5m / s, and the relative distance is 3m. The calculated I=0.89. The Kalman filter predicts I=0.85 after 0.5 seconds. Historical data shows that the average traffic index in this area is 0.82. The current scene complexity is medium, and the weight coefficient =0.5, =0.3, =0.2, final P=0.86.

[0049] S105 : Determine the target sector space with the highest traffic index as the avoidance direction, and select a corresponding target obstacle avoidance solution according to the channel width parameter of the target sector space.

[0050] After calculating the accessibility index for each of the 16 sectors, the AGV robot control system first sorts the accessibility index in descending order and selects the sector with the highest accessibility index as the target direction. The system then fits the obstacle boundary using LiDAR point cloud data and calculates the channel width of the target sector. The obstacle avoidance level is categorized as follows: when the channel width is ≥2 times the vehicle width, deceleration is used for obstacle avoidance; when it is between 1.5 and 2 times the vehicle width, lane change is used for obstacle avoidance; and when it is less than 1.5 times the vehicle width, emergency avoidance is triggered.

[0051] After the initial obstacle avoidance plan is determined, the system monitors the obstacle avoidance effect parameters (relative distance change rate, heading deviation, obstacle avoidance progress) in real time. When the obstacle avoidance effect is detected to be below the threshold, the plan upgrade mechanism is activated. The obstacle avoidance margin index is calculated based on the remaining decision time and maneuvering space. The obstacle avoidance margin index is calculated as: ,in, is the remaining decision time, v is the current speed, w is the channel width, The minimum safe width.

[0052] S106. When the hardware control layer is triggered, emergency braking is immediately performed, and after emergency braking, driving is resumed according to the control instructions issued by the main control system.

[0053] When the risk level exceeds 0.9, triggering the hardware control layer, the system directly controls the motor braking system via PWM signals. The braking system utilizes a dual-circuit redundant design, comprising electromagnetic brakes and mechanical brakes. During braking, the system monitors wheel speed signals in real time and uses the anti-lock braking system (ABS) to prevent wheel lock.

[0054] After an emergency brake, the AGV robot enters a safe stop, cutting off power output and maintaining the brakes. The system sends an emergency brake report to the main control system via the 5G communication module, containing information such as the location, time, and reason for the brake. After assessing the situation, the main control system issues a recovery command, which the AGV receives, releasing the brakes and replanning its path.

[0055] It should be noted that the main control system establishes a wireless connection with the remote collaboration layer of the AGV robot control system of each vehicle, receives obstacle information uploaded by each AGV robot, and generates control instructions for all AGV robots within a preset range nearby based on the obstacle information and sends them to the AGV robot control system of the corresponding AGV robot, thereby realizing obstacle avoidance control.

[0056] S107. When the remote collaboration layer is triggered, the AGV robot is controlled to wait for control instructions issued by the main control system during normal driving, and the AGV robot is controlled to avoid obstacles according to the control instructions.

[0057] When the risk level falls below 0.7, triggering the remote collaboration layer, the AGV robot maintains its current path while simultaneously uploading environmental data to the main control system via the 5G communication module. Based on global path planning, the main control system generates optimal obstacle avoidance instructions. While waiting for instructions, the AGV robot uses a predictive control algorithm to maintain a safe speed, ensuring sufficient maneuvering space when receiving instructions.

[0058] Upon receiving a command from the master control, the AGV robot executes a lane change or deceleration maneuver using a fuzzy PID controller. The system transmits the execution status in real time, and the master control system adjusts subsequent commands based on this feedback. If no command is received within a preset time, the AGV automatically downgrades to the local computing layer for processing.

[0059] In the above-mentioned embodiment, the AGV robot control system achieves hierarchical rapid response for the AGV robot by establishing a three-level obstacle avoidance mechanism consisting of a hardware control layer, a local computing layer, and a remote collaboration layer. The system automatically triggers the most appropriate obstacle avoidance level based on the risk level, eliminating the need to wait for instructions from the master control system each time. When the risk is low, the local computing layer can independently make obstacle avoidance decisions, quickly selecting the optimal obstacle avoidance path through 16-sector spatial analysis and traffic index evaluation. This hierarchical control strategy reduces obstacle avoidance response time and avoids the processing delays caused by traditional solutions where all obstacle avoidance decisions require the master control system to participate, effectively reducing the risk of collision caused by waiting for the master control system to process.

[0060] The following is a more detailed description of the process of the method provided by this implementation. Figure 2 , is another flow chart of a hierarchically controlled AGV robot obstacle avoidance method in an embodiment of the present application.

[0061] S201. When triggering the local computing layer according to the risk level value, the motion characteristic coefficient of the AGV robot is calculated according to the speed parameter of the AGV robot.

[0062] After triggering the local computation layer, the AGV control system first evaluates the AGV's dynamic motion characteristics to determine its obstacle avoidance capabilities in different directions. Specifically, the AGV collects its current velocity (v), friction coefficient (μ), and acceleration (g) in real time via onboard sensors. These values ​​are then substituted into the kinematic coefficient K, calculating the AGV's kinematic coefficient using the formula.

[0063] S202 : Divide the obstacle avoidance area within the preset range into a front area, a side area, and a rear area according to the motion characteristic coefficient.

[0064] Based on the motion characteristic coefficient K, the AGV robot control system divides the obstacle avoidance area within a preset range (usually 10 meters in front, 5 meters to the side, and 3 meters to the rear) into three areas: forward, side, and rear, and further subdivides them into 16 sector-shaped spaces.

[0065] The forward area, centered on the AGV's current direction of travel, extends within a ±180° range and is divided into eight 45° sectors. Because this area encompasses the AGV's primary travel path, the highest obstacle detection accuracy is required. The lateral area extends ±90° on each side, with each side divided into three 60° sectors, for a total of six sectors. The sector division angles in the lateral area are larger than those in the forward area to balance obstacle detection range and computational complexity. The rearward area, centered on the AGV's tail, extends within a ±45° range and is divided into two 45° sectors. Because the AGV rarely reverses, the rearward area has the fewest sectors.

[0066] Optionally, when the K value increases, the sector division angle of the forward area will decrease (for example, from 45° to 30°) to improve the resolution of obstacles in front; when the K value decreases, the number of sectors in the side area can be increased to 8 to enhance the perception of side obstacles. This is not limited here.

[0067] S203. Calculate a spatial weight coefficient based on the sector-shaped spatial distribution, and use the spatial weight coefficient as a weighting parameter in a traffic index calculation formula.

[0068] Specifically, since the forward direction is the main driving direction of the AGV robot and the obstacle threat level is the highest, it is assigned the highest weight, with a weight coefficient of 0.4 for the forward area. Similarly, the weight coefficients for the lateral area and the rearward area are 0.3, respectively.

[0069] Optionally, when the AGV robot is performing special tasks (such as carrying extra-long cargo), the system can automatically increase the weight of the lateral area to 0.4 and reduce the weight of the forward area to 0.35 to prioritize cargo transportation safety. In addition, if a moving obstacle (such as a person following) is detected in the rear direction, the rearward weight can be temporarily increased to 0.4.

[0070] S204: Collect the real-time status parameters of each sector space, and calculate the instant pass index based on the real-time status parameters.

[0071] After completing the sector-shaped space division, the AGV robot control system uses multi-source sensors to collect parameters such as obstacle density, moving speed and relative distance of each sector space in real time, and calculates the instant pass index based on these data.

[0072] The obstacle density (ρ) is determined by counting the ratio of the number of obstacle point clouds in the sector space to the volume of the space. For example, if the volume of a sector space in the forward direction is 10m³ and 5 obstacle points are detected, then ρ = 0.5. The obstacle's velocity is calculated by the displacement difference between two consecutive frames of point cloud data, and the relative velocity is calculated by combining it with the AGV's own velocity. The relative distance (d) is the closest distance to the obstacle measured by the lidar. The formula for calculating the instant pass index is not detailed here.

[0073] S205. Based on the changing trend of the real-time state parameters in the historical sampling period, a Kalman filter algorithm is used to predict the state evolution of each sector space within a preset time window to obtain a short-term traffic index.

[0074] The AGV robot control system uses the real-time state parameters of the historical sampling period (usually 100ms) to predict the obstacle state evolution within a preset time window (such as 0.5 seconds) in the future through the Kalman filter algorithm, and substitutes the predicted obstacle density, relative speed and distance into the instantaneous pass index formula to obtain the short-term pass index S.

[0075] S206. Establish a long-term evaluation model based on the historical traffic records of each sector space, and calculate the long-term traffic index corresponding to each sector space.

[0076] The AGV robot control system analyzes the historical traffic records of each sector to establish a long-term evaluation model and calculate the long-term traffic index (L). The local database records data such as the number of traffic, average traffic time, and number of accidents in each sector over the past 24 hours. The long-term evaluation model uses the exponentially weighted moving average (EWMA) algorithm to calculate the long-term traffic index. The specific formula is: , where a is the forgetting factor (usually 0.1), is the instant access index at the current moment, is the long-term traffic index at the previous moment.

[0077] Optionally, when it is detected that a sector space has failed to pass through three times in a row, the system will automatically multiply the long-term pass index of the area by a penalty coefficient of 0.8; if it is successfully passed through five times in a row, it will be multiplied by a reward coefficient of 1.2.

[0078] S207 , adjusting the weight coefficients corresponding to the instantaneous pass index, the short-term pass index, and the long-term pass index respectively according to the complexity of the current scenario, and obtaining the pass index through weighted combination.

[0079] The AGV control system dynamically adjusts the weighting coefficients of the immediate access index (I), short-term access index (S), and long-term access index (L) based on the current scene complexity, and calculates the final access index (P) through a weighted combination. Scene complexity is determined by factors such as obstacle type (fixed / mobile), density, and AGV speed. For example, a scene is considered high complexity if the detected percentage of mobile obstacles exceeds 50% and the AGV speed is greater than 1.5 m / s.

[0080] Furthermore, the weight coefficient can be dynamically adjusted as follows: High complexity scenario: =0.6 (Instant Pass Index Coefficient), =0.3 (short-term traffic index coefficient), =0.1 (long-term traffic index coefficient); medium complexity scenario: =0.5, =0.3, =0.2; low complexity scenario: =0.4, =0.2, =0.4.

[0081] S208. After confirming the avoidance direction, calculate the obstacle avoidance level according to the channel width parameter of the target sector space, and determine the initial obstacle avoidance plan according to the obstacle avoidance level.

[0082] After determining the target sector space, the AGV robot control system fits the obstacle boundary through the lidar point cloud data, calculates the channel width, and selects the initial obstacle avoidance plan based on preset standards.

[0083] Specifically, after cluster analysis of the LiDAR point cloud data, the minimum enclosing rectangle method is used to calculate the effective channel width (w) of the target sector. Initial strategies are then implemented based on the obstacle avoidance level. For deceleration and obstacle avoidance, w ≥ 2 vehicle widths (e.g., for a 0.8-meter vehicle width, w ≥ 1.6 meters); for lane change and obstacle avoidance, w ≤ 1.5 vehicle widths < 2 vehicle widths; and for emergency obstacle avoidance, w < 1.5 vehicle widths.

[0084] Optionally, in deceleration and obstacle avoidance, the system reduces the AGV robot speed to below 1m / s through a PID controller; in lane change and obstacle avoidance, the system plans an S-shaped path with a maximum steering angle limit of 30°; in emergency obstacle avoidance, the emergency braking system of the hardware control layer is triggered.

[0085] S209: After the initial obstacle avoidance plan is started, if it is detected that the obstacle avoidance effect parameter of the AGV robot is lower than the preset target threshold, the plan upgrade judgment is triggered.

[0086] When executing the initial obstacle avoidance plan, the AGV robot control system monitors the obstacle avoidance effect parameters in real time. When the parameters fall below the preset threshold, the plan upgrade is triggered.

[0087] Specifically, the relative distance change rate must be ≥0.5m / s, the heading deviation tolerance must be ±15°, and the obstacle avoidance progress must be updated every 100ms, with a requirement of ≥10% / cycle. During threshold comparison, if the relative distance change rate is <0.5m / s and the heading deviation tolerance is >15°, or the obstacle avoidance progress is <5% / cycle, the system will determine that the obstacle avoidance is ineffective and trigger the solution upgrade process.

[0088] The AGV robot control system of the embodiment of the present invention is an electronic device. Figure 3 A schematic diagram of the architecture of an electronic device suitable for implementing an embodiment of the present invention is shown.

[0089] It should be noted that Figure 3 The electronic device shown is only an example and should not limit the functions and scope of use of the embodiments of the present invention.

[0090] Those skilled in the art will appreciate that all or part of the steps in the various methods of the above embodiments can be accomplished by instructions (computer programs) or by controlling related hardware through instructions (computer programs), and the instructions can be stored in a computer-readable storage medium and loaded and executed by a processor. The electronic device of this embodiment includes a storage medium and a processor, wherein the storage medium stores a plurality of instructions, which can be loaded by the processor to execute any step of the method provided in the embodiment of the present invention.

[0091] Specifically, the storage medium and the processor are electrically connected directly or indirectly to achieve data transmission or interaction. For example, these elements can be electrically connected to each other via one or more signal lines. The storage medium stores computer-executable instructions for implementing the data access control method, including at least one software function module that can be stored in the storage medium in the form of software or firmware. The processor executes various functional applications and data processing by running the software programs and modules stored in the storage medium. The storage medium can be, but is not limited to, random access memory (RAM), read-only memory (ROM), programmable read-only memory (PROM), erasable programmable read-only memory (EPROM), electrically erasable programmable read-only memory (EEPROM), etc. The storage medium is used to store programs, and the processor executes the programs after receiving the execution instructions.

[0092] Furthermore, the software programs and modules in the above-mentioned storage medium may also include an operating system, which may include various software components and / or drivers for managing system tasks (such as memory management, storage device control, power management, etc.), and may communicate with various hardware or software components to provide an operating environment for other software components. The processor may be an integrated circuit chip having signal processing capabilities. The above-mentioned processor may be a general-purpose processor, including a central processing unit (CPU), a network processor (NP), etc., which may implement or execute the various methods, steps, and logic flow diagrams disclosed in this embodiment. The general-purpose processor may be a microprocessor or the processor may also be any conventional processor, etc.

[0093] Since the instructions stored in the storage medium can execute the steps of any method provided in the embodiments of the present invention, the beneficial effects of any method provided in the embodiments of the present invention can be achieved. Please refer to the previous embodiments for details and will not be repeated here.

[0094] The above description is merely a preferred embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any changes or substitutions that can be easily conceived by a person skilled in the art within the technical scope disclosed in the present invention should be included in the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be based on the scope of protection of the claims.

Claims

1. A hierarchically controlled AGV robot obstacle avoidance method, applied to an AGV robot control system, wherein the AGV robot control system includes three obstacle avoidance actuators: a hardware control layer, a local computing layer, and a remote collaboration layer, characterized in that: The method comprises: Collecting AGV robot operating parameters, including load value, ground friction coefficient, obstacle distance value, historical accident data and mission time parameters; Inputting the operating parameters into a risk assessment model and outputting a risk level value; The obstacle avoidance actuator to be triggered first is determined based on the preset interval in which the risk level value is located. The preset interval includes a first preset interval corresponding to the hardware control layer, a second preset interval of the local computing layer, and a third preset interval of the remote collaboration layer. The hardware control layer is used to perform emergency braking on the AGV robot, and the remote collaboration layer is connected to the main control system. When the local computing layer is triggered, the obstacle avoidance area of ​​the AGV robot within the preset range is divided into 16 sector spaces, and the traffic index of each sector space is calculated according to the traffic index calculation formula; The target sector space with the highest traffic index is determined as the avoidance direction, and a corresponding target obstacle avoidance solution is selected according to the channel width parameter of the target sector space.

2. The method according to claim 1, characterized in that The step of inputting the operating parameters into the risk assessment model and outputting the risk level value specifically includes: The hardware control layer performs a threshold judgment based on the load value, the ground friction coefficient, and the obstacle distance value to obtain a first risk index; The local computing layer inputs the first risk index, the historical accident data, and the task time parameter into a fuzzy evaluator, and calculates a second risk index based on a preset fuzzy rule; The remote collaboration layer inputs the first risk index and the second risk index into a pre-trained deep learning model to obtain a third risk index; The risk level value is determined according to a weighted average of the first risk index, the second risk index, and the third risk index.

3. The method according to claim 1, characterized in that The step of dividing the obstacle avoidance area of ​​the AGV robot within a preset range into 16 sector-shaped spaces specifically includes: Calculating the motion characteristic coefficient of the AGV robot based on the speed parameters of the AGV robot, wherein the speed parameters include current speed, steering angle, and acceleration; Dividing the obstacle avoidance area within the preset range into a forward area, a side area, and a rearward area according to the motion characteristic coefficient, wherein the forward area, the side area, and the rearward area correspond to 8 sector spaces, 6 sector spaces, and 2 sector spaces, respectively; The spatial weight coefficient is calculated according to the sector-shaped spatial distribution, and the spatial weight coefficient is used as a weighting parameter in the traffic index calculation formula.

4. The method according to claim 1, wherein The step of calculating the traffic index of each sector space according to the traffic index calculation formula specifically includes: Collecting real-time status parameters of each sector space and calculating an instantaneous traffic index based on the real-time status parameters, wherein the real-time status parameters include obstacle density, moving speed, and relative distance; Based on the changing trend of the real-time state parameters in the historical sampling period, a Kalman filter algorithm is used to predict the state evolution of each sector space within a preset time window to obtain a short-term traffic index; Establishing a long-term evaluation model based on the historical traffic records of each sector space and calculating the long-term traffic index corresponding to each sector space; The weight coefficients corresponding to the instantaneous traffic index, the short-term traffic index and the long-term traffic index are adjusted respectively according to the complexity of the current scene, and the traffic index is obtained through weighted combination.

5. The method according to claim 1, wherein The step of selecting a corresponding target obstacle avoidance solution according to the channel width parameter of the target sector space specifically includes: Calculating an obstacle avoidance level based on the channel width parameter of the target sector space, and determining an initial obstacle avoidance plan based on the obstacle avoidance level, wherein the obstacle avoidance levels are deceleration obstacle avoidance, lane change obstacle avoidance, and emergency obstacle avoidance from low to high; After the initial obstacle avoidance plan is started, if the obstacle avoidance effect parameters of the AGV robot are detected to be lower than the preset target threshold, the plan upgrade judgment is triggered. The obstacle avoidance effect parameters include relative distance change rate, heading deviation and obstacle avoidance progress; In the scheme upgrade judgment, when it is detected that the obstacle avoidance margin index is greater than a preset upgrade threshold, the initial obstacle avoidance scheme is upgraded to a target obstacle avoidance scheme corresponding to a higher obstacle avoidance level. The obstacle avoidance margin index is determined based on the remaining decision time and maneuvering space.

6. The method according to claim 1, wherein After the step of determining the obstacle avoidance actuator to be triggered first according to the preset interval of the risk level value, the method further includes: When the hardware control layer is triggered, emergency braking is immediately performed; After the emergency braking, the vehicle resumes driving according to the control instructions issued by the main control system.

7. The method according to claim 1, characterized in that After the step of determining the obstacle avoidance actuator to be triggered first according to the preset interval of the risk level value, the method further includes: When the remote collaboration layer is triggered, the AGV robot is controlled to wait for the control instructions issued by the main control system during normal driving; The AGV robot is controlled to avoid obstacles according to the control instructions.

8. An AGV robot control system, characterized in that: The AGV robot control system includes: one or more processors and memories; The memory is coupled to the one or more processors, and the memory is used to store computer program code, where the computer program code includes computer instructions. The one or more processors call the computer instructions to enable the AGV robot control system to execute the method according to any one of claims 1 to 7.

9. A computer-readable storage medium comprising instructions, characterized in that: When the instruction is executed on the AGV robot control system, the AGV robot control system is caused to execute the method according to any one of claims 1 to 7.

10. A computer program product, characterized in that When the computer program product is run on an AGV robot control system, the AGV robot control system is caused to execute the method according to any one of claims 1 to 7.

Images