An automatic safety obstacle avoidance system and method for forklift navigation

Abstract

The present invention discloses an automatic safety obstacle avoidance system and method for forklift navigation, which relates to the field of forklift safety technology. The system includes the following components: a sensor module, a navigation module, a virtual barrier setting module and a control module. The present invention significantly improves the adaptability of the forklift in complex environments through an intelligent path planning algorithm and a dynamic obstacle avoidance decision algorithm. The control module adopts a dynamic obstacle avoidance decision algorithm based on genetic algorithm optimization, which can quickly and accurately evaluate the obstacle avoidance strategy and ensure that the forklift can make the best avoidance action when approaching an obstacle.

Description

Technical Field

[0001] The present invention relates to the technical field of forklift safety, and in particular to an automatic safety obstacle avoidance system and method for forklift navigation. Background Art

[0002] In the field of logistics and warehousing, forklifts are core handling equipment, and their navigation and obstacle avoidance technologies are of great significance to improving operational efficiency and ensuring operational safety. With the rapid development of intelligent manufacturing and automation technologies, forklift navigation systems are gradually transforming and upgrading towards intelligence and autonomy, aiming to achieve more efficient and safer material handling operations.

[0003] There are deficiencies in traditional technologies. When planning paths, traditional control modules cannot fully consider environmental complexity, obstacle shape and material factors, resulting in a less than optimal planned path or even a risk of collision. At the same time, obstacle avoidance strategies mostly use fixed safety distances and obstacle avoidance actions, lack the ability to dynamically respond to environmental changes, and are unable to make the best obstacle avoidance decisions based on actual conditions.

[0004] In summary, the traditional forklift navigation system is difficult to meet the high requirements of the modern logistics and warehousing industries for the intelligence and automation of forklift navigation systems. Therefore, it is particularly important to develop an automatic safety obstacle avoidance system and method for forklift navigation. Summary of the invention

[0005] The purpose of the present invention is to make up for the deficiencies of the prior art and to provide an automatic safe obstacle avoidance system and method for forklift navigation. The system significantly improves the navigation and obstacle avoidance capabilities of the forklift in complex environments through an intelligent path planning algorithm and a dynamic obstacle avoidance decision algorithm.

[0006] In order to solve the above technical problems, the present invention provides the following technical solutions: an automatic safety obstacle avoidance system for forklift navigation, the system comprising a sensor module, a navigation module, a virtual barrier setting module and a control module;

[0007] The sensor module is composed of multi-dimensional sensing units, which can not only detect distance, but also sense the shape and material characteristics of objects. Each sensing unit converts the collected information into a specific data format and transmits it to the control module.

[0008] The navigation module uses an intelligent path planning algorithm based on the current position coordinates (X c Y c Z c ), target An automatic and safe obstacle avoidance method for forklift navigation Position coordinates (X t Y t Z t) and the surrounding environment information set E, the path planning is performed by the following formula: Path = f(P c P t E) = λ × S (P c P t )+μ×C(E), where S(P c P t ) is a spatial vector function based on the current position and target position of the forklift, which is obtained by calculating the angle and distance between the forklift posture vector and the target vector. The specific formula is: Among them, θ is the angle between the forklift posture vector and the target vector, and C(E) is the environment complexity evaluation function, which is obtained by weighted summation, that is, Where n is the number of influencing factors, w i is the weight coefficient of each factor, F i (E) is the evaluation function corresponding to each factor, and the weight coefficient w i The source is obtained through deep learning training of different types of forklift operation scene data, so that path planning can adapt to various complex environments;

[0009] The virtual barrier setting module allows the operator to set virtual obstacles and restricted areas through a visual interactive interface, and the setting information is stored in the system's cache database;

[0010] The control module: When the forklift approaches the virtual barrier, it receives information from the sensor module and the navigation module, and makes a decision through a dynamic obstacle avoidance decision algorithm. The algorithm formula is: Decision = g(D sensor ,V barrier ,P forklift )=ρ×D sensor +σ×V barrier +τ×P forklift An automatic and safe obstacle avoidance method for forklift navigation, wherein D sensor Represents the multidimensional information set of the forklift and obstacles detected by the sensor, V barrier Represents the attribute parameter set of the virtual screen, P forklift represents the current state parameter set of the forklift, and the weight coefficients ρ, σ, and τ are determined by reinforcement learning training on a large number of actual operation data of forklifts in different environments.

[0011] Furthermore, the multidimensional perception unit in the sensor module adopts a combination of quantum sensing technology and traditional sensing technology to achieve high-precision and high-sensitivity environmental perception. The quantum sensing part assists in judging the material and shape of surrounding objects by detecting tiny changes in physical quantities. Each multidimensional perception unit has independent calibration parameters, which are automatically adjusted when the system starts to adapt to different environmental conditions. The calibration parameters are determined by performing multiple measurements on known standard objects and continuously adjusting the parameters using an optimization algorithm to minimize the error between the measurement results and the actual values.

[0012] Furthermore, the multidimensional sensing unit in the sensor module adopts a combination of quantum sensing technology and traditional sensing technology. The quantum sensing part assists in judging the material and shape of surrounding objects by detecting tiny changes in physical quantities. Each multidimensional sensing unit has independent calibration parameters, which are automatically adjusted when the system starts to adapt to different environmental conditions. The calibration parameters are determined by performing multiple measurements on known standard objects and continuously adjusting the parameters using an optimization algorithm to minimize the error between the measurement results and the actual values.

[0013] Furthermore, the environment complexity evaluation function C(E) in the navigation module determines the weight coefficient w of each factor. i When the forklift is running, the adaptive weight adjustment algorithm is used to make a preliminary assessment of each influencing factor and assign an initial weight according to its potential impact on the forklift's driving. During the operation of the forklift, data is continuously collected according to the actual situation, and the weight of each factor is dynamically adjusted. The specific formula for weight adjustment is: in, is the adjusted weight, is the weight before adjustment, Δw i is the weight adjustment amount, which is obtained by analyzing and calculating the actual operation data and is expressed as Δw i =ζ×I factor , where ζ is the adjustment factor, I factor Influencing factor indicators.

[0014] Furthermore, the virtual barrier setting module supports dynamic virtual barrier setting, which automatically adjusts the position and properties of the virtual barrier according to the real-time operating status of the forklift and changes in the surrounding environment. When the forklift is carrying specific goods, the virtual barrier range is automatically set according to the size and weight of the goods. The dynamic virtual barrier is set by real-time monitoring of the forklift's position, speed, and cargo information parameters, and using a fuzzy logic algorithm for judgment and adjustment. The input parameters in the fuzzy logic algorithm include the distance between the forklift and surrounding objects, the size of the cargo, and the speed of the forklift, and the output is the adjustment parameters of the virtual barrier.

[0015] Furthermore, when determining the weight coefficients ρ, σ, and τ, the dynamic obstacle avoidance decision algorithm in the control module adopts an optimization method based on a genetic algorithm to randomly generate a set of initial weight coefficients, and evaluates the obstacle avoidance effect corresponding to each set of weight coefficients by simulating the operation of the forklift in different environments. Based on the evaluation results, the weight coefficients are selected for crossover and mutation operations to generate new weight coefficient combinations, and this process is repeated until the optimal weight coefficient combination is found.

[0016] Furthermore, the system also includes an intelligent early warning module, which analyzes the data from the sensor module and the navigation module, predicts potential dangerous situations in advance, and issues early warning signals. The triggering conditions of the early warning signals are determined by logical judgment, and a deep learning algorithm is used to train historical data to improve the accuracy and timeliness of the early warning.

[0017] Furthermore, the system also includes a data fusion module, which fuses the multi-dimensional data collected by the sensor module to improve the accuracy and reliability of the data. The data fusion adopts a multi-sensor fusion algorithm, comprehensively considers the measurement accuracy and reliability factors of different sensors, and performs data fusion through weighted averaging and Kalman filtering methods. The weight is determined by performing performance evaluation on each sensor and dynamically adjusting it according to the evaluation results.

[0018] On the other hand, an automatic safety obstacle avoidance method for forklift navigation is characterized in that the method comprises the following specific steps:

[0019] S100, sensor: It is composed of multi-dimensional sensing units, which can not only detect distance, but also sense the shape and material characteristics of objects. Each sensing unit converts the collected information into a specific data format and transmits it to the control module;

[0020] S200, navigation module: uses intelligent path planning algorithm, based on the current position coordinates of the forklift (X c Y c Z c ), target position coordinates (X t Y t Z t ) and the surrounding environment information set E, the path planning is performed by the following formula: Path = f(P c P t E) = λ × S (P c P t )+μ×C(E), where S(P c P t ) is a spatial vector function based on the current position and target position of the forklift, λ and μ are weight coefficients, which are obtained by calculating the angle and distance between the forklift posture vector and the target vector. The specific formula is: Among them, θ is the angle between the forklift posture vector and the target vector, and C(E) is the environment complexity evaluation function, which is obtained by weighted summation, that is, Where n is the number of influencing factors, w i is the weight coefficient of each factor, F i (E) is the evaluation function corresponding to each factor, and the weight coefficient w i The source is obtained through deep learning training of a large number of different types of forklift operation scene data, so that path planning can adapt to various complex environments;

[0021] S300, virtual barrier setting module: allows the operator to set virtual obstacles and restricted areas through a visual interactive interface, and the setting information is stored in the system's cache database;

[0022] S400, control module: When the forklift approaches the virtual barrier, it receives information from the sensor module and the navigation module, and makes a decision through the dynamic obstacle avoidance decision algorithm. The algorithm formula is: Decision = g(D sensor ,V barrier ,P forklift )=ρ×D sensor +σ×V barrier +τ×P forklift , where D sensor Represents the multidimensional information set of the forklift and obstacles detected by the sensor, V barriet Represents the attribute parameter set of the virtual screen, P forklift represents the current state parameter set of the forklift, and the weight coefficients ρ, σ, and τ are determined by reinforcement learning training on a large number of actual operation data of forklifts in different environments.

[0023] Compared with the prior art, an automatic safety obstacle avoidance system and method for forklift navigation has the following advantages:

[0024] Beneficial effects:

[0025] 1. The system significantly improves the forklift's adaptability in complex environments through intelligent path planning algorithms and dynamic obstacle avoidance decision algorithms. The control module adopts a dynamic obstacle avoidance decision algorithm optimized based on genetic algorithms, which can quickly and accurately evaluate obstacle avoidance strategies to ensure that the forklift can make the best avoidance action when approaching obstacles.

[0026] 2. By integrating high-precision, multi-dimensional sensor modules, the system can perceive the shape, material and distance information of surrounding objects in real time, providing detailed and accurate environmental data for the navigation module. Combined with intelligent path planning algorithms and adaptive weight adjustment mechanisms, the system can quickly and reasonably plan paths for environments of different complexities, effectively avoid collisions, and ensure the stability and safety of the forklift during transportation. BRIEF DESCRIPTION OF THE DRAWINGS

[0027] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the prior art descriptions are briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention, and for ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0028] Figure 1 An operational flow chart of an automatic safety obstacle avoidance system and method for forklift navigation. DETAILED DESCRIPTION

[0029] The technical solutions in the embodiments of the present invention are described clearly and completely below. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of them. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0030] Embodiment 1

[0031] This embodiment describes that in a large-scale storage and logistics center, forklifts, as key handling equipment, need to operate efficiently and safely in a complex and changeable cargo stacking environment. The internal layout of the warehouse is intricate, and the cargo is of various types with different shapes, sizes, and materials. At the same time, there are other forklifts, personnel, and automated equipment running synchronously, which places extremely high demands on the forklift's navigation and obstacle avoidance capabilities.

[0032] The sensor module equipped on the forklift integrates multiple high-precision multi-dimensional sensing units. These units combine the advantages of quantum sensing technology and traditional sensing technology, and can detect the distance, shape and material characteristics of surrounding objects in real time and accurately. When the forklift approaches goods with irregular shapes or special materials, the sensor module can quickly identify the contour and hardness of the goods. Key information, and convert this information into a specific data format and transmit it to the control module, providing data support for subsequent path planning and obstacle avoidance decisions.

[0033] The navigation module calculates the optimal path based on the forklift's current position coordinates, target position coordinates, and surrounding environment information using an intelligent path planning algorithm. c Y c Z c ), target position coordinates (X t Y t Z t ) and the surrounding environment information set E, the path planning is performed by the following formula: Path = f(P c Pt E) = λ × S (P c P t )+μ×C(E), where S(P c P t ) is a spatial vector function based on the current position and target position of the forklift, λ and μ are weight coefficients, which are obtained by calculating the angle and distance between the forklift posture vector and the target vector. The specific formula is: Among them, θ is the angle between the forklift posture vector and the target vector, and C(E) is the environment complexity evaluation function, which is obtained by weighted summation, that is, Where n is the number of influencing factors, w i is the weight coefficient of each factor, F i (E) is the evaluation function corresponding to each factor. The environmental complexity evaluation function plays a key role in the path planning process. It will perform a weighted summation of multiple influencing factors and dynamically adjust the weights according to the actual situation to ensure that the path planning can adapt to various complex environments. By adopting an adaptive weight adjustment algorithm, the system can continuously collect data according to the actual situation and dynamically adjust the weights of each factor, making the path planning more reasonable and efficient.

[0034] Operators can set up virtual barriers in the system through a visual interactive interface to restrict forklifts from traveling in certain areas or near sensitive goods. When a forklift is carrying fragile or heavy goods, the system will automatically adjust the position and properties of the virtual barrier, such as setting a wider buffer zone or a higher obstacle height limit, to ensure that the forklift does not damage the goods or cause other safety hazards due to misoperation during the handling process. The dynamic virtual barrier setting function can monitor the forklift's position, speed, and cargo information parameters in real time, and use fuzzy logic algorithms to make judgments and adjustments, making the setting of the virtual barrier more flexible and intelligent.

[0035] The control module is the core part of the system, which is responsible for receiving information from the sensor module and the navigation module, and making judgments and decisions through the dynamic obstacle avoidance decision algorithm. sensor ,V barrier ,P forklift )=ρ×D sensor +σ×V barrier +τ×P forklift , where D sensor Represents the multidimensional information set of the forklift and obstacles detected by the sensor, V barrier Represents the attribute parameter set of the virtual screen, P forkliftrepresents the current state parameter set of the forklift. The weight coefficients ρ, σ, and τ are determined by reinforcement learning training on a large number of actual operation data of forklifts in different environments. The dynamic obstacle avoidance decision algorithm adopts an optimization method based on a genetic algorithm. By simulating the operation of the forklift in different environments, the obstacle avoidance effect corresponding to each set of weight coefficients is evaluated, and finally the optimal combination of weight coefficients is found. When the forklift approaches a virtual barrier or an actual obstacle, the control module can quickly and accurately evaluate the obstacle avoidance strategy and issue corresponding control instructions to ensure that the forklift can make the best avoidance action.

[0036] It is also equipped with an intelligent early warning module, which analyzes the data from the sensor module and navigation module, predicts potential dangerous situations in advance, and issues early warning signals. The triggering conditions of the early warning signals are determined through logical judgment, and a deep learning algorithm is used to train historical data to improve the accuracy and timeliness of the early warning. The data fusion module is responsible for fusing the multi-dimensional data collected by the sensor module to improve the accuracy and reliability of the data. By using multi-sensor fusion algorithms and weighted averaging and Kalman filtering methods for data fusion, the system can perceive the surrounding environment information more comprehensively and provide more accurate data support for path planning and obstacle avoidance decisions.

[0037] Embodiment 2

[0038] This embodiment describes that at a port terminal, a container handling forklift is responsible for the important task of unloading containers from a ship and transporting them to a storage yard or transporting them from a storage yard to a ship. The port environment is complex and changeable, including large ships, busy loading and unloading equipment, storage yards, container stacks, and a large number of staff. The forklift needs to frequently avoid obstacles during driving to ensure the safe transportation of goods and avoid collision accidents and delays.

[0039] Deploy high-precision, high-reliability multi-dimensional sensing units, including quantum sensing technology combined with traditional sensing technology, to accurately detect the distance, shape and material characteristics of surrounding objects. These sensing units can capture dynamic changes in the port in real time, including berthing and unberthing of ships, movement of loading and unloading equipment, stacking and movement of containers, and walking of staff, ensuring that forklifts can accurately perceive the surrounding environment. The sensor module has environmental adaptability and can work normally in severe weather and complex lighting conditions.

[0040] Using intelligent path planning algorithm, based on the current position coordinates (X c Y c Z c ), target position coordinates (X t Y t Z t ) and the surrounding environment information set E, the path planning is performed by the following formula: Path = f(P c Pt E) = λ × S (P c P t )+μ×C(E), where S(P c P t ) is a spatial vector function based on the current position and target position of the forklift, λ and μ are weight coefficients, which are obtained by calculating the angle and distance between the forklift posture vector and the target vector. The specific formula is: Among them, θ is the angle between the forklift posture vector and the target vector, and C(E) is the environment complexity evaluation function, which is obtained by weighted summation, that is, Where n is the number of influencing factors, w i is the weight coefficient of each factor, F i (E) is the evaluation function corresponding to each factor, which ensures that the path planning can adapt to the complex environment of the port. The system can update the path planning in real time to cope with sudden traffic congestion and equipment failure in the port, ensuring that the forklift can reach the destination efficiently and safely.

[0041] Through the visual interactive interface, operators can intuitively view the port layout, forklift positions, and obstacle distribution information. According to the port layout and container handling requirements, operators can set virtual obstacles and restricted areas at key locations, such as loading and unloading areas, around ships, and yard entrances to prevent forklifts from accidentally entering. The system supports dynamic virtual barrier settings and automatically adjusts the position and properties of virtual barriers according to changes in the ship's berthing position and the movement of loading and unloading equipment.

[0042] When the forklift approaches the virtual barrier, the control module receives information from the sensor module and the navigation module, and makes a judgment through the dynamic obstacle avoidance decision algorithm. The algorithm comprehensively considers the multi-dimensional information set of the forklift and the obstacle detected by the sensor, the attribute parameter set of the virtual barrier, and the current state parameter set of the forklift. Decision = g(D sensor ,V barrier ,P forklift )=ρ×D sensor +σ×V barrier +τ×P forklift An automatic and safe obstacle avoidance method for forklift navigation, wherein D sensor Represents the multidimensional information set of the forklift and obstacles detected by the sensor, V barrier Represents the attribute parameter set of the virtual screen, P forkliftIt represents the current state parameter set of the forklift. The weight coefficients ρ, σ, and τ are determined by reinforcement learning training on a large number of actual operation data of forklifts in different environments. The optimal combination of weight coefficients is determined by the optimization method based on genetic algorithm to ensure that the forklift can safely avoid obstacles. In extreme cases, such as when the forklift is about to collide with an obstacle, the system can automatically trigger the emergency braking function to ensure that the forklift stops in time to avoid collision accidents.

[0043] The automatic safety obstacle avoidance system is integrated with other management systems of the port terminal to achieve data sharing and collaborative work. The system can transmit the forklift's operating status and obstacle avoidance information to other management systems of the port terminal in real time, so that managers can conduct real-time monitoring and scheduling.

[0044] It will be apparent to those skilled in the art that the invention is not limited to the details of the exemplary embodiments described above and that the invention can be implemented in other specific forms without departing from the spirit or essential features of the invention. Therefore, the embodiments should be considered exemplary and non-limiting in all respects, and the scope of the invention is defined by the appended claims rather than the foregoing description, and it is intended that all variations falling within the meaning and scope of the equivalent elements of the claims be included in the invention. Any reference numeral in a claim should not be considered as limiting the claim to which it relates.

Claims

1. An automatic safety obstacle avoidance system for forklift navigation, characterized in that: The system includes a sensor module, a navigation module, a virtual barrier setting module and a control module; The sensor module is composed of multi-dimensional sensing units, which can not only detect distance, but also sense the shape and material characteristics of objects. Each sensing unit converts the collected information into a specific data format and transmits it to the control module. The navigation module uses an intelligent path planning algorithm based on the current position coordinates (X c Y c Z c ), target position coordinates (X t Y t Z t ) and the surrounding environment information set E, the path planning is performed by the following formula: Path = f(P c P t E) = λ × S (P c P t )+μ×C(E), where S(P c P t ) is a spatial vector function based on the current position and target position of the forklift, λ and μ are weight coefficients, which are obtained by calculating the angle and distance between the forklift posture vector and the target vector. The specific formula is: Among them, θ is the angle between the forklift posture vector and the target vector, and C(E) is the environment complexity evaluation function, which is obtained by weighted summation, that is, Where n is the number of influencing factors, w i is the weight coefficient of each factor, F i (E) is the evaluation function corresponding to each factor, and the weight coefficient w i The source is obtained through deep learning training of different types of forklift operation scene data, so that path planning can adapt to various complex environments; The virtual barrier setting module allows the operator to set virtual obstacles and restricted areas through a visual interactive interface, and the setting information is stored in the system's cache database; The control module: When the forklift approaches the virtual barrier, it receives information from the sensor module and the navigation module, and makes a decision through a dynamic obstacle avoidance decision algorithm. The algorithm formula is: Decision = g(D sensor ,V barrier ,P forklift )=ρ×D sensor +σ×V barrier +τ×P forklift , where D sensor Represents the multidimensional information set of the forklift and obstacles detected by the sensor, V barrier Represents the attribute parameter set of the virtual screen, P forklift represents the current state parameter set of the forklift, and the weight coefficients ρ, σ, and τ are determined by reinforcement learning training on a large number of actual operation data of forklifts in different environments.

2. The automatic safety obstacle avoidance system for forklift navigation according to claim 1, characterized in that: The multidimensional sensing unit in the sensor module adopts a combination of quantum sensing technology and traditional sensing technology. The quantum sensing part assists in judging the material and shape of surrounding objects by detecting tiny changes in physical quantities. Each multidimensional sensing unit has independent calibration parameters, which are automatically adjusted when the system starts to adapt to different environmental conditions. The calibration parameters are determined by measuring known standard objects multiple times and continuously adjusting the parameters using an optimization algorithm to minimize the error between the measurement result and the actual value.

3. The automatic safety obstacle avoidance system for forklift navigation according to claim 1, characterized in that: The environment complexity evaluation function C(E) in the navigation module determines the weight coefficient w of each factor. i When the forklift is running, the adaptive weight adjustment algorithm is used to make a preliminary assessment of each influencing factor and assign an initial weight according to its potential impact on the forklift's driving. During the operation of the forklift, data is continuously collected according to the actual situation, and the weight of each factor is dynamically adjusted. The specific formula for weight adjustment is: in, is the adjusted weight, is the weight before adjustment, Δw i is the weight adjustment amount, which is obtained by analyzing and calculating the actual operation data and is expressed as Δw i =ζ×I factor , where ζ is the adjustment factor, I factor Influencing factor indicators.

4. The automatic safety obstacle avoidance system for forklift navigation according to claim 1, characterized in that: The virtual barrier setting module supports dynamic virtual barrier setting, automatically adjusts the position and attributes of the virtual barrier according to the real-time operating status of the forklift and changes in the surrounding environment, and automatically sets the virtual barrier range according to the size and weight of the cargo when the forklift is carrying specific cargo. The dynamic virtual barrier is set by real-time monitoring of the forklift's position, speed, and cargo information parameters, and using a fuzzy logic algorithm for judgment and adjustment. The input parameters in the fuzzy logic algorithm include the distance between the forklift and surrounding objects, cargo size, and forklift speed, and the output is the adjustment parameters of the virtual barrier.

5. The automatic safety obstacle avoidance system for forklift navigation according to claim 1, characterized in that: The dynamic obstacle avoidance decision algorithm in the control module adopts an optimization method based on a genetic algorithm to randomly generate a set of initial weight coefficients when determining the weight coefficients ρ, σ, and τ. By simulating the operation of a forklift in different environments, the obstacle avoidance effect corresponding to each set of weight coefficients is evaluated. According to the evaluation results, the weight coefficients are selected for crossover and mutation operations to generate new weight coefficient combinations. This process is repeated until the optimal weight coefficient combination is found.

6. The automatic safety obstacle avoidance system for forklift navigation according to claim 1, characterized in that: The system also includes an intelligent early warning module, which analyzes data from the sensor module and the navigation module to predict potential dangerous situations in advance and issue early warning signals. The triggering conditions of the early warning signals are determined through logical judgment, and a deep learning algorithm is used to train historical data to improve the accuracy and timeliness of the early warning.

7. The automatic safety obstacle avoidance system for forklift navigation according to claim 1, characterized in that: The system also includes a data fusion module, which fuses the multi-dimensional data collected by the sensor module to improve the accuracy and reliability of the data. The data fusion adopts a multi-sensor fusion algorithm, comprehensively considers the measurement accuracy and reliability factors of different sensors, and performs data fusion through weighted averaging and Kalman filtering methods. The weight is determined by performing performance evaluation on each sensor and dynamically adjusting it according to the evaluation results.

8. A real-time laser coal discing method, characterized in that: The specific steps of this method are: S100, sensor: It is composed of multi-dimensional sensing units, which can not only detect distance, but also sense the shape and material characteristics of objects. Each sensing unit converts the collected information into a specific data format and transmits it to the control module; S200, navigation module: uses intelligent path planning algorithm, based on the current position coordinates of the forklift (X c Y c Z c ), target position coordinates (X t Y t Z t ) and the surrounding environment information set E, the path planning is performed by the following formula: Path = f(P c P t E) = λ × S (P c P t )+μ×C(E), where S(P c P t ) is a spatial vector function based on the current position and target position of the forklift, which is obtained by calculating the angle and distance between the forklift posture vector and the target vector. The specific formula is: Among them, θ is the angle between the forklift posture vector and the target vector, and C(E) is the environment complexity evaluation function, which is obtained by weighted summation, that is, Where n is the number of influencing factors, w i is the weight coefficient of each factor, F i (E) is the evaluation function corresponding to each factor, and the weight coefficient w i The source is obtained through deep learning training of a large number of different types of forklift operation scene data, so that path planning can adapt to various complex environments; S300, virtual barrier setting module: allows the operator to set virtual obstacles and restricted areas through a visual interactive interface, and the setting information is stored in the system's cache database; S400, control module: When the forklift approaches the virtual barrier, it receives information from the sensor module and the navigation module, and makes a decision through the dynamic obstacle avoidance decision algorithm. The algorithm formula is: Decision = g(D sensor ,V barrier ,P forklift )=ρ×D sensor +σ×V barrier +τ×P forklift , where D sensor Represents the multidimensional information set of the forklift and obstacles detected by the sensor, V barrier Represents the attribute parameter set of the virtual screen, P forklift represents the current state parameter set of the forklift, and the weight coefficients ρ, σ, and τ are determined by reinforcement learning training on a large number of actual operation data of forklifts in different environments.

Images