A method and system for preventing collisions of the reclaiming arm of a stacker-reclaimer.

Abstract

This invention provides a collision avoidance method and system for the reclaimer arm of a stacker-reclaimer, relating to the field of data processing technology. The method includes acquiring first, second, and third information; performing three-dimensional modeling based on the first information to obtain a three-dimensional model; simulating the stacker-reclaimer's operating scenario based on the second, third, and three-dimensional model to obtain the simulation results; obtaining fourth information; and obtaining a target collision avoidance scheme. This invention, through operating scenario simulation and risk simulation processing of the stacker-reclaimer, can pre-simulate operational conditions, accurately identify potential collision risks, effectively reduce the probability of collisions between the reclaimer arm and surrounding equipment, and utilizes a deep learning model for multi-objective optimization. The resulting target collision avoidance scheme not only avoids operational interruptions caused by frequent collision avoidance or accidental collisions but also enables the stacker-reclaimer to complete reclaiming tasks more smoothly and efficiently, reducing waiting time and operational adjustments, and improving operational efficiency.

Description

Technical Field

[0001] This invention relates to the field of data processing technology, and more specifically, to a method and system for preventing collisions with the reclaiming arm of a stacker-reclaimer. Background Technology

[0002] In modern bulk material yards, multiple stacker-reclaimers need to operate in coordination. Due to the large range of motion of the stacker-reclaimer's boom, there is a significant risk of collisions between stacker-reclaimers operating nearby, between stacker-reclaimers and fixed structures, and between stacker-reclaimers and material piles. A collision can cause serious machine damage and personal injury, resulting in substantial economic losses for the company. Traditional stacker-reclaimer collision prevention control relies primarily on operator visual observation, supplemented by anti-collision limit switches, wire rope limits, and microwave radar switches. However, this method depends heavily on the operator's skill level and the sensitivity and accuracy of the sensors, resulting in relatively low accuracy. Summary of the Invention

[0003] The purpose of this invention is to provide a method and system for preventing collisions with the reclaimer arm of a stacker-reclaimer, in order to improve the aforementioned problems. To achieve the above objective, the technical solution adopted by this invention is as follows:

[0004] In a first aspect, this application provides a method for preventing collisions with the reclaiming arm of a stacker-reclaimer, including:

[0005] Acquire first information, second information and third information. The first information includes the structure and parameter information of the stacker-reclaimer. The second information includes the site layout of the stacker-reclaimer's operating area, the distribution of surrounding equipment and the operating trajectory information. The third information includes the historical operating data of the stacker-reclaimer.

[0006] Based on the first information, a three-dimensional modeling process is performed to obtain a three-dimensional model;

[0007] Based on the second information, the third information, and the three-dimensional model, the operation scenario of the stacker-reclaimer is simulated, and the simulation results are obtained.

[0008] The simulation results of the operation scenario are subjected to risk simulation processing to obtain fourth information, which includes the distribution of peripheral devices and the corresponding collision situation under different operation scenarios.

[0009] Based on a pre-defined deep learning model, the fourth information is subjected to multi-objective optimization processing to obtain a target collision avoidance scheme.

[0010] Secondly, this application also provides a collision avoidance system for the reclaimer arm of a stacker-reclaimer, comprising:

[0011] The acquisition unit is used to acquire first information, second information and third information. The first information includes the structure and parameter information of the stacker-reclaimer. The second information includes the site layout, distribution of surrounding equipment and running trajectory information of the stacker-reclaimer's working area. The third information includes the historical operating data of the stacker-reclaimer.

[0012] The first modeling unit is used to perform three-dimensional modeling processing based on the first information to obtain a three-dimensional model.

[0013] The first simulation unit is used to simulate the operation scenario of the stacker-reclaimer based on the second information, the third information and the three-dimensional model, and obtain the operation scenario simulation results;

[0014] The simulation unit is used to perform risk simulation processing on the simulation results of the operation scenario to obtain fourth information, which includes the distribution of peripheral devices and the corresponding collision situation under different operation scenarios.

[0015] The first optimization unit is used to perform multi-objective optimization processing on the fourth information based on a preset deep learning model to obtain a target collision avoidance scheme.

[0016] The beneficial effects of this invention are as follows:

[0017] This invention simulates the operation scenarios and risks of the stacker-reclaimer, enabling it to preview operational conditions, accurately identify potential collision risks, effectively reduce the probability of collisions between the reclaiming arm and surrounding equipment, and utilizes a deep learning model for multi-objective optimization to generate a target collision avoidance scheme. This not only prevents operational interruptions caused by frequent collision avoidance or accidental collisions, but also allows the stacker-reclaimer to complete material reclaiming tasks more smoothly and efficiently, reducing unnecessary waiting time and operational adjustments, and significantly improving operational efficiency.

[0018] Other features and advantages of the invention will be set forth in the description which follows, and will be apparent in part from the description, or may be learned by practicing embodiments of the invention. Attached Figure Description

[0019] To more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings used in the embodiments will be briefly introduced below. It should be understood that the following drawings only show some embodiments of the present invention and should not be regarded as a limitation on the scope. For those skilled in the art, other related drawings can be obtained based on these drawings without creative effort.

[0020] Figure 1 This is a schematic diagram of the anti-collision method for the stacker-reclaimer arm described in an embodiment of the present invention;

[0021] Figure 2This is a schematic diagram of the anti-collision system of the stacker-reclaimer arm in an embodiment of the present invention.

[0022] The diagram is labeled as follows: 10, acquisition unit; 20, first modeling unit; 30, first simulation unit; 40, simulation unit; 50, first optimization unit. Detailed Implementation

[0023] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of the present invention. The components of the embodiments of the present invention described and shown in the accompanying drawings can generally be arranged and designed in various different configurations. Therefore, the following detailed description of the embodiments of the present invention provided in the accompanying drawings is not intended to limit the scope of the claimed invention, but merely to illustrate selected embodiments of the invention. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without inventive effort are within the scope of protection of the present invention.

[0024] It should be noted that similar reference numerals and letters in the following figures indicate similar items; therefore, once an item is defined in one figure, it does not need to be further defined and explained in subsequent figures. Furthermore, in the description of this invention, terms such as "first," "second," etc., are used only to distinguish descriptions and should not be construed as indicating or implying relative importance.

[0025] Example 1:

[0026] This embodiment provides a collision prevention method for the reclaiming arm of a stacker-reclaimer.

[0027] See Figure 1 The figure shows that the method includes steps S10, S20, S30, S40 and S50.

[0028] Step S10. Obtain first information, second information and third information. The first information includes the structure and parameter information of the stacker-reclaimer. The second information includes the site layout, distribution of surrounding equipment and running trajectory information of the stacker-reclaimer's working area. The third information includes the historical operating data of the stacker-reclaimer.

[0029] Specifically, the first information includes the structure and parameter information of the stacker-reclaimer, covering the cantilever length, bucket wheel size, connection method of each mechanism, operating speed, slewing angle range, cantilever lifting stroke, etc.; the second information includes the site layout, distribution of surrounding equipment, and operating trajectory information of the stacker-reclaimer's operating site, including the location of the material pile, track direction, and location of surrounding buildings, the distribution of surrounding equipment including the parking and activity areas of other stacker-reclaimers, transport vehicles, etc. in the same area, and the operating trajectory information including the regular operating routes of other equipment in the operating site; the third information includes the historical operating data of the stacker-reclaimer, including past operating time, operating process, frequency of action of each mechanism, and the time, location, degree of equipment damage caused by the collision, and cause of the collision.

[0030] Step S20. Perform 3D modeling processing based on the first information to obtain a 3D model;

[0031] Specifically, step S20 includes steps S21 to S25:

[0032] Step S21. Perform parametric modeling based on the first information to obtain the initial 3D model;

[0033] Specifically, the structure and parameter information of the stacker-reclaimer are presented in an intuitive three-dimensional form, resulting in a static three-dimensional model of the stacker-reclaimer.

[0034] Step S22. Based on adaptive mesh generation technology, perform mesh generation on the initial 3D model to obtain the finite element model;

[0035] Specifically, adaptive meshing technology is used to mesh the initial 3D model to obtain a finite element model, which can more accurately simulate the actual structure and mechanical properties of the stacker-reclaimer. By rationally meshing and refining the mesh in key areas, the computational accuracy is improved, making subsequent analysis results more accurate and reliable.

[0036] Step S23. Perform modal analysis and static loading simulation analysis on the finite element model to obtain the corresponding dynamic and static characteristics, respectively;

[0037] Specifically, modal analysis and static loading simulation analysis are performed on the finite element model. Modal analysis can reveal the response of the stacker-reclaimer at different vibration frequencies, helping to avoid problems such as resonance. Static loading simulation analysis can reveal the deformation and stress distribution of the stacker-reclaimer under different loads, providing a basis for structural strength assessment.

[0038] Step S24. Based on dynamic characteristics, static characteristics, the operating speed and braking performance of the stacker-reclaimer, determine the sensitivity data of the stacker-reclaimer;

[0039] Specifically, a set of design parameters closely related to the performance of the stacker-reclaimer should be defined, including but not limited to boom length, turning radius, drive motor power, and braking system parameters. These parameters directly affect the stacker-reclaimer's operating speed, stability, and material handling efficiency. Numerical simulation software or analysis tools should be used to adjust these parameters one by one within a reasonable design parameter range. For example, the boom length can be gradually increased or decreased, while recording the stacker-reclaimer's operating status and performance data for each parameter change.

[0040] Monitor and record the impact of each parameter change on the stacker-reclaimer's performance indicators, which cover multiple aspects such as reclaiming efficiency, operational stability, structural stress distribution, and energy consumption. For example, observe the changes in reclaiming efficiency and energy consumption when the drive motor power is changed; monitor changes in operational stability and structural stress when the slewing radius is adjusted.

[0041] A comprehensive assessment of the impact of each design parameter on various performance indicators is conducted, and the assessment results are presented in the form of sensitivity data, also known as key information. Through this sensitivity data, it is possible to clearly determine which design parameters have the most significant impact on the performance indicators of the stacker-reclaimer.

[0042] These sensitivity data provide crucial references for the optimized design of stacker-reclaimers. For example, if it is found that the boom length has a significant impact on the material handling efficiency, subsequent design optimizations can focus on adjusting the boom length parameter to improve the overall performance of the stacker-reclaimer, including increasing material handling efficiency, enhancing operational stability, and reducing energy consumption.

[0043] Step S25. Based on the finite element model, dynamic characteristics, static characteristics, and sensitivity data, construct a three-dimensional model of the stacker-reclaimer;

[0044] Specifically, a three-dimensional model is constructed based on the finite element model, dynamic characteristics, static characteristics, and sensitivity data. This integrates information from multiple aspects, enabling the constructed three-dimensional model to more comprehensively and accurately reflect the actual operating status and performance characteristics of the stacker-reclaimer, providing more reliable model support for subsequent collision avoidance analysis and scheme formulation.

[0045] Step S30. Based on the second information, the third information, and the three-dimensional model, simulate the operation scenario of the stacker-reclaimer to obtain the simulation results.

[0046] Specifically, using professional simulation software, the operating space of the stacker-reclaimer is set according to the site layout. Potential interference factors and possible collision points are identified based on the distribution of surrounding equipment. Historical operating data is referenced to set the operating parameters of the stacker-reclaimer under different working conditions, such as speed and acceleration. Through comprehensive simulation, the final simulation results of the operating scenario are obtained, which can intuitively present the operating status of the stacker-reclaimer under various possible real-world conditions.

[0047] Specifically, step S30 includes steps S31 to S34:

[0048] Step S31. Based on the preset peripheral devices of different types, specifications and operating states, and the corresponding preset operating rules, perform device simulation processing on the second information to obtain the basic model of device simulation;

[0049] Specifically, by using probability and statistics-based models, the operating modes of different equipment in the site can be simulated, such as the start-up and shutdown time, operation duration, and slewing angle distribution probability of stacker-reclaimers, as well as the probability distribution of the driving routes of surrounding transport vehicles and the statistical patterns of waiting time for loading and unloading, which helps to more realistically reflect the actual operating conditions.

[0050] Step S32. Analyze and process the third information to obtain the equipment operation mode and potential collision risk points;

[0051] Specifically, this step utilizes pattern recognition technology to analyze the stacker-reclaimer's operational data. This technology can extract key patterns and features from massive amounts of operational data, helping to understand the stacker-reclaimer's operating modes and identify potential failure risk points. Further, a method combining deep learning algorithms with traditional graph theory and complex network theory is employed: deep learning algorithms can automatically mine complex patterns and correlations in the data, viewing the stacker-reclaimer's operating state as a changing pattern in time and space dimensions, thereby accurately identifying operating modes and performance patterns under different working conditions. Simultaneously, using graph theory and complex network theory, an operational network model of the stacker-reclaimer is constructed to deeply study the connection relationships between components, operating path characteristics, and potential jamming or stress concentration points. These analytical methods work synergistically to help comprehensively understand the stacker-reclaimer's operational structure and dynamic changes, providing strong support for discovering potential equipment failure hazards and safety risks, and providing a reliable basis for optimizing equipment operation and formulating maintenance strategies.

[0052] Step S33. Combine the basic equipment simulation model, equipment operation mode and potential collision risk points for predictive processing to obtain abnormal simulation results under equipment failure and abnormal operation line conditions;

[0053] Specifically, probabilistic models are used to simulate the changing operating states of stacker-reclaimers under different working conditions. These conditions may include changes in material bulk density, variations in stacker-reclaimer operating speed, and the impact of sudden malfunctions on stacker-reclaimer operation. Through probabilistic model simulation, the dynamic changes in stacker-reclaimer operation under different conditions can be predicted, providing more accurate input conditions for stacker-reclaimer operation and control.

[0054] Furthermore, fault tree analysis was employed to analyze and assess various factors that could lead to stacker-reclaimer malfunctions. These factors encompass equipment failures, such as motor damage and hydraulic system leaks; human error, such as incorrect instructions or violations of operating procedures; and uncontrollable factors, such as severe weather and sudden power outages. By constructing a fault tree, the causal relationships between different events can be clearly identified, and the potential impact of each factor on the stacker-reclaimer's operation can be assessed.

[0055] This step comprehensively considers the simulation results under various operational failure scenarios, including material blockage, enabling a more comprehensive and in-depth understanding of the actual operating conditions of the stacker-reclaimer in complex working environments. This provides strong support for developing more effective maintenance strategies, optimizing operating processes, and improving the reliability and stability of the equipment.

[0056] Step S34. Based on the 3D model and the results of the abnormal simulation, perform optimization processing, combine the structural dimensions, operating speed and braking performance of the stacker-reclaimer and surrounding equipment, predict the response of the sensors and the operation adjustment strategy made by the stacker-reclaimer, and obtain the simulation results of the operation scenario.

[0057] Specifically, it comprehensively considers all actual parameters and abnormal conditions of the equipment, and can accurately predict the operation adjustment strategy and sensor response of the stacker-reclaimer in complex operating scenarios, providing precise guidance for optimizing the operation control of the stacker-reclaimer and ensuring that the stacker-reclaimer is safe and efficient in actual operation.

[0058] Step S40. Perform risk simulation processing on the simulation results of the running scenario to obtain the fourth information, which includes the distribution of peripheral devices and the corresponding collision situation under different running scenarios;

[0059] Specifically, step S40 includes steps S41 to S44:

[0060] Step S41. Analyze the simulation results of the stacker-reclaimer's operation scenarios to obtain a set of scenarios for different peripheral equipment and corresponding operating lines;

[0061] Specifically, based on the simulation results of the stacker-reclaimer's operation scenarios, a set of scenarios is obtained, including different peripheral equipment and their corresponding operating routes. In this step, a thorough analysis of the large amount of data generated from previous simulations is conducted, categorizing and organizing the different types of peripheral equipment (such as transport vehicles, other loading and unloading equipment, etc.) and their respective operating routes when working collaboratively with the stacker-reclaimer, forming specific set of scenarios. The benefit of this approach is that it systematically organizes complex operational situations, providing a clear data foundation for subsequent precise analysis of the interaction between the stacker-reclaimer and peripheral equipment, and facilitating the identification of potential operational conflicts and safety hazards.

[0062] Step S42. Identify the scene set based on the preset hidden Markov model to obtain the stacker-reclaimer operation mode identification result;

[0063] Specifically, Hidden Markov Models (HMMs) are powerful statistical models capable of handling data containing hidden states. In this step, this model is used to analyze a set of scenarios, abstracting the operational behavior of the stacker-reclaimer (CRR) under different scenarios into different operating modes. For example, it identifies the different material handling modes of the CCR when there is sufficient material and when there is a shortage of material, as well as the coordination modes when surrounding equipment is busy and idle. By accurately identifying these operating modes, we can better understand the operating patterns of the CCR, providing a basis for predicting its future operating status and developing targeted control strategies.

[0064] Step S43. Based on the operation mode recognition results of the stacker-reclaimer, construct a collaborative operation simulation model by treating the stacker-reclaimer and peripheral equipment as independent entities and simulating their interaction behavior.

[0065] Specifically, after clarifying the operating mode of the stacker-reclaimer, the stacker-reclaimer and its peripheral equipment are considered as individuals with independent operating logic but interconnected. Computer simulation technology is used to simulate their interactive behaviors in actual operation, such as material transfer and obstacle avoidance along the operating path. A virtual simulation environment is constructed in space, and the stacker-reclaimer and peripheral equipment representing the individuals are placed within it. The stacker-reclaimer and peripheral equipment will interact and move according to preset operating rules and modes. The constructed collaborative operation simulation model can intuitively demonstrate the collaborative operation process between the equipment.

[0066] Step S44. Run the simulation engine and analyze the collaborative simulation model of the operation, and extract the average number of braking times, collision prediction probability and safe running time ratio to obtain the fourth information;

[0067] Specifically, by running the simulation engine, the collaborative simulation model of the operation runs in the virtual environment according to the set rules. Then, the data during the operation is analyzed and processed. Among them, the average number of braking can reflect the frequency of start-stop of the stacker-reclaimer during operation, which indirectly reflects the smoothness of operation; the collision prediction probability can intuitively show the probability of the stacker-reclaimer colliding with surrounding equipment; and the safe running time percentage can measure the safe and stable operation of the stacker-reclaimer throughout the entire operation process.

[0068] Step S50. Based on the preset deep learning model, perform multi-objective optimization processing on the fourth information to obtain the target collision avoidance scheme;

[0069] Specifically, step S50 includes steps S51 to S53:

[0070] Step S51. Based on the fourth information and the preset Pareto front optimization model, model the process for maximizing safety, maximizing operation efficiency and minimizing energy consumption to obtain at least two initial collision avoidance schemes.

[0071] Specifically, a fourth set of information, including average braking frequency, collision prediction probability, and percentage of safe operating time, is input into a pre-defined Pareto front optimization model. This model aims to maximize safety, maximize operational efficiency, and minimize energy consumption. During the modeling process, the operating parameters and performance indicators of the stacker-reclaimer and surrounding equipment are fully considered. Through complex mathematical calculations and optimization algorithms, the relationships between different objectives are weighed, ultimately generating at least two initial collision avoidance schemes. These initial schemes represent preliminary results under multi-objective optimization, each with its own emphasis, providing a foundation for further selection and optimization. This multi-objective optimization model comprehensively considers multiple key factors in the operation of the stacker-reclaimer, avoiding the limitations of pursuing only a single objective (such as focusing solely on safety while neglecting operational efficiency or energy consumption).

[0072] Step S52. Simulate and evaluate the performance of the initial collision avoidance scheme under various operating scenarios based on the genetic algorithm to obtain the optimal collision avoidance scheme set;

[0073] Specifically, step S52 includes steps S521 to S525:

[0074] Step S521. Perform population initialization processing according to the preliminary anti-collision scheme to obtain the initial population, and set the initial population under different running scenarios. Each individual in the initial population represents an anti-collision scheme.

[0075] Step S522. Calculate the fitness of all individuals in the initial population in the same running scenario to obtain the fitness of multiple individuals;

[0076] Step S523. Perform selection, crossover, and mutation operations based on individual fitness values ​​to obtain the updated population;

[0077] Step S524. Perform fitness evaluation and convergence test on the updated population, and iterate to obtain the final target population;

[0078] Step S525. Determine the optimal individual and the corresponding optimal anti-collision scheme from multiple final target populations corresponding to different operating scenarios, and form an optimal anti-collision scheme set;

[0079] Specifically, a genetic algorithm is used to simulate and evaluate the generated initial collision avoidance schemes. The genetic algorithm simulates the process of biological evolution, treating each initial collision avoidance scheme as an "individual." Through operations such as selection, crossover, and mutation, it simulates the performance of the schemes under various operating scenarios. During the simulation, each scheme is scored according to set evaluation indicators (such as the number of collisions, job completion time, and energy consumption). After multiple rounds of iteration, the best-performing schemes are selected to form the optimal collision avoidance scheme set. The schemes in this set exhibit relatively good overall performance under various operating scenarios.

[0080] Genetic algorithms can quickly and efficiently search for superior solutions from a large number of possible schemes. By simulating various operating scenarios, the feasibility and effectiveness of the solutions are comprehensively evaluated. The resulting set of optimal collision avoidance schemes, which gathers multiple high-performance solutions, provides high-quality candidate schemes for the final determination of the target collision avoidance scheme.

[0081] Step S53. Process the set of optimal collision avoidance schemes based on the preset deep learning, and perform policy selection and dynamic adjustment by setting a reward mechanism and policy gradient method to obtain the target collision avoidance scheme;

[0082] Specifically, step S53 includes steps S531 to S536:

[0083] Step S531. Based on the preset deep learning model, extract features from the set of optimal collision avoidance schemes to obtain a set of feature vectors;

[0084] Specifically, deep learning models, with their complex internal neural network structures such as convolutional layers and fully connected layers, perform in-depth mining and analysis of the data in each collision avoidance scheme. This data includes the stacker-reclaimer's operating parameters (such as speed, acceleration, and rotation angle), surrounding environmental information (such as material stacking patterns and the location of surrounding equipment), and time series information. Through the model's calculations, each scheme is transformed into a corresponding feature vector, and numerous feature vectors constitute a feature vector set.

[0085] Step S532. Analyze the feature vector set to determine the key dimension information related to collision avoidance, and construct a mathematical model mapping the key dimension information and the preset reward standard;

[0086] Specifically, through data analysis methods and professional knowledge, key dimensions directly related to collision avoidance are selected from numerous feature dimensions, such as time to collision (TTC), minimum safe distance, and relative velocity change rate. Then, using these key dimensions as independent variables and a pre-defined reward standard (typically determined based on collision avoidance effectiveness, operational efficiency, energy consumption, and other factors) as the dependent variable, a mathematical model is constructed to accurately describe the relationship between the two. This model can be a linear regression model, a nonlinear function model, etc., depending on the intrinsic relationship between the key dimensions and the reward standard; no particular restrictions are imposed here.

[0087] Step S533. Initialize the policy space and calculate the reward value of the candidate policies contained in the policy space based on the deep learning model and the mapping mathematical model;

[0088] Specifically, the strategy space is first initialized, containing candidate strategies with various combinations of control parameters, such as the stacker-reclaimer's steering control parameters, speed regulation parameters, and braking control parameters. Then, a pre-defined deep learning model is used to analyze the candidate strategies. Simultaneously, a constructed mapping mathematical model is used, inputting the feature vectors corresponding to the candidate strategies into the deep learning model to obtain relevant feature information. This information is then substituted into the mapping mathematical model to calculate the reward value for each candidate strategy under the current settings. The reward value reflects the candidate strategy's comprehensive performance in terms of collision avoidance, operational efficiency, and energy consumption.

[0089] By calculating the reward value of candidate policies, a quantitative evaluation of numerous candidate policies in the policy space can be achieved. This helps to quickly screen out the best-performing policies, avoid blind attempts, save optimization time and costs, and provide clear direction for subsequent parameter tuning based on policy gradient algorithms.

[0090] Step S534. Calculate the gradient degree of each policy parameter based on the policy gradient algorithm, and adjust the parameters of the candidate policy according to the preset update rule to obtain the adjusted candidate policy;

[0091] Specifically, the policy gradient algorithm calculates the gradient value for each policy parameter by differentiating the log-likelihood function of the policy with respect to the policy parameters and incorporating the reward value. These gradient values ​​reflect the direction and extent of the impact of parameter changes on the reward value. Then, according to a pre-defined update rule, the calculated gradient values ​​are used to adjust the parameters of the candidate policy. The update rule is typically based on the principle of gradient ascent or descent, adjusting the parameters in the direction that maximizes the reward value, thus obtaining the adjusted candidate policy.

[0092] Step S535. Adjust the policy parameters according to the gradient value and the preset update rule to obtain the updated candidate policy;

[0093] Step S536. Calculate and perform convergence verification on the adjusted candidate strategy reward value, and iterate to obtain the optimal candidate strategy, i.e., the target collision avoidance scheme;

[0094] Specifically, the adjusted candidate strategy is re-input into the preset deep learning model and mapping mathematical model. The deep learning model analyzes the new strategy features, and the mapping mathematical model recalculates the reward value of the candidate strategy based on the new feature information. The reward value of the candidate strategy after each adjustment is continuously calculated, and the trend of the reward value is observed. When the change in the reward value is less than a preset range, the strategy is considered to have converged, that is, reached a relatively stable and optimal state. At this point, the corresponding candidate strategy is the optimal candidate strategy obtained through multiple iterations, and it is determined as the target collision avoidance scheme. The final target collision avoidance scheme is optimal under the current optimization framework and constraints, effectively reducing collision risk, ensuring the safe, efficient, and energy-saving operation of equipment, and improving the enterprise's production efficiency and safety.

[0095] Example 2:

[0096] like Figure 2 As shown, this embodiment provides a collision avoidance system for the reclaimer arm of a stacker-reclaimer. The system includes:

[0097] The acquisition unit 10 is used to acquire first information, second information and third information. The first information includes the structure and parameter information of the stacker-reclaimer, the second information includes the site layout, distribution of surrounding equipment and running trajectory information of the stacker-reclaimer's working area, and the third information includes the historical operating data of the stacker-reclaimer.

[0098] The first modeling unit 20 is used to perform three-dimensional modeling processing based on the first information to obtain a three-dimensional model.

[0099] The first simulation unit 30 is used to simulate the operation scenario of the stacker-reclaimer based on the second information, the third information, and the three-dimensional model, and to obtain the simulation results of the operation scenario.

[0100] Simulation unit 40 is used to perform risk simulation processing on the simulation results of the running scenario to obtain fourth information, which includes the distribution of peripheral devices and the corresponding collision situation under different running scenarios.

[0101] The first optimization unit 50 is used to perform multi-objective optimization processing on the fourth information based on a preset deep learning model to obtain a target collision avoidance scheme.

[0102] In one specific embodiment disclosed in this application, the first simulation unit 30 includes:

[0103] The second simulation unit is used to perform equipment simulation processing on the second information based on the preset peripheral equipment of different types, specifications and operating states, as well as the corresponding preset operating rules, to obtain the basic model of equipment simulation.

[0104] The first analysis unit is used to analyze and process the third information to obtain the equipment operation mode and potential collision risk points;

[0105] The predictive processing unit is used to combine the basic model of equipment simulation, equipment operation mode and potential collision risk points to perform predictive processing and obtain abnormal simulation results under equipment failure and abnormal operation line conditions.

[0106] The second optimization unit is used to perform optimization processing based on the three-dimensional model and abnormal simulation results. It combines the structural dimensions, operating speed and braking performance of the stacker-reclaimer and surrounding equipment to predict the response of the sensors and the operation adjustment strategies made by the stacker-reclaimer, and obtain the simulation results of the operation scenario.

[0107] In one specific embodiment disclosed in this application, the first modeling unit 20 includes:

[0108] The second modeling unit is used to perform parametric modeling processing based on the first information to obtain the initial three-dimensional model;

[0109] The meshing element is used to perform meshing processing on the initial three-dimensional model based on adaptive meshing technology to obtain the finite element model.

[0110] The second simulation unit is used to perform modal analysis and static loading simulation analysis on the finite element model, and obtain the corresponding dynamic and static characteristics respectively.

[0111] The first determining unit is used to determine the sensitivity data of the stacker-reclaimer based on dynamic characteristics, static characteristics, the operating speed and braking performance of the stacker-reclaimer.

[0112] The first building unit is used to construct a three-dimensional model of the stacker-reclaimer based on the finite element model, dynamic characteristics, static characteristics, and sensitivity data.

[0113] In one specific embodiment disclosed in this application, the simulation unit 40 includes:

[0114] The second analysis unit is used to analyze the simulation results of the stacker-reclaimer's operation scenarios to obtain a set of scenarios for different peripheral equipment and corresponding operating lines.

[0115] The identification unit is used to identify the scene set based on the preset hidden Markov model to obtain the identification result of the stacker-reclaimer operation mode;

[0116] The second building unit is used to construct a collaborative simulation of operations based on the operation mode recognition results of the stacker-reclaimer. It treats the stacker-reclaimer and peripheral equipment as independent individuals and simulates their interaction behavior to obtain a collaborative simulation model of operations.

[0117] The third analysis unit is used to run and analyze the collaborative simulation model of the operation, and extract the average number of braking times, collision prediction probability and safe running time ratio to obtain the fourth information.

[0118] In one specific embodiment disclosed in this application, the first optimization unit 50 includes:

[0119] The third modeling unit is used to perform modeling processing based on the fourth information and the preset Pareto front optimization model, aiming to maximize safety, maximize operation efficiency and minimize energy consumption, and obtain at least two initial collision avoidance schemes.

[0120] The evaluation unit is used to simulate and evaluate the performance of the initial collision avoidance scheme under various operating scenarios based on the genetic algorithm, and obtain the optimal collision avoidance scheme set.

[0121] The processing unit is used to process the set of optimal collision avoidance schemes based on a preset deep learning method. By setting a reward mechanism and a policy gradient method, it performs policy selection and dynamic adjustment to obtain the target collision avoidance scheme.

[0122] In one specific embodiment disclosed in this application, the evaluation unit includes:

[0123] The first initialization unit is used to perform population initialization processing according to the preliminary anti-collision scheme, obtain the initial population, and set the initial population under different operating scenarios. Each individual in the initial population represents an anti-collision scheme.

[0124] The first calculation unit is used to calculate the fitness of all individuals in the initial population in the same running scenario, and obtain the fitness of multiple individuals.

[0125] Mutation units are used to obtain an updated population based on individual fitness values ​​through selection, crossover, and mutation operations.

[0126] The first verification unit is used to evaluate the fitness and verify the convergence of the updated population, and iteratively obtain the final target population.

[0127] The constituent units are used to determine the optimal individual and the corresponding optimal anti-collision scheme from multiple final target populations corresponding to different operating scenarios, and form an optimal anti-collision scheme set.

[0128] In one specific embodiment disclosed in this application, the processing unit includes:

[0129] The extraction unit is used to extract features from the set of optimal anti-collision schemes based on a preset deep learning model to obtain a set of feature vectors.

[0130] The parsing unit is used to parse the feature vector set, determine the key dimension information related to collision avoidance, and construct a mathematical model mapping the key dimension information and the preset reward standard.

[0131] The second initialization unit is used to initialize the policy space and calculate the reward value of the candidate policies contained in the policy space based on the deep learning model and the mapping mathematical model.

[0132] The second calculation unit is used to calculate the gradient degree of each policy parameter based on the policy gradient algorithm, and adjust the parameters of the candidate policy according to the preset update rule to obtain the adjusted candidate policy.

[0133] The adjustment unit is used to adjust the policy parameters based on the gradient value according to a preset update rule to obtain the updated candidate policy.

[0134] The second verification unit is used to calculate the reward value of the adjusted candidate strategy and perform convergence verification, and iteratively obtain the optimal candidate strategy as the target collision avoidance scheme.

[0135] It should be noted that the specific methods by which each module performs operations in the system described in the above embodiments have been described in detail in the embodiments related to the method, and will not be elaborated here.

[0136] The above description is merely a preferred embodiment of the present invention and is not intended to limit the invention. Various modifications and variations can be made to the present invention by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

[0137] The above description is merely a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the technical scope disclosed in the present invention should be included within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the scope of the claims.

Claims

1. A method for preventing collisions with the reclaiming arm of a stacker-reclaimer, characterized in that, include: Acquire first information, second information and third information. The first information includes the structure and parameter information of the stacker-reclaimer. The second information includes the site layout of the stacker-reclaimer's operating area, the distribution of surrounding equipment and the operating trajectory information. The third information includes the historical operating data of the stacker-reclaimer. Based on the first information, a three-dimensional modeling process is performed to obtain a three-dimensional model; Based on the second information, the third information, and the three-dimensional model, the operation scenario of the stacker-reclaimer is simulated, and the simulation results are obtained. The simulation results of the operation scenario are subjected to risk simulation processing to obtain fourth information, which includes the distribution of peripheral devices and the corresponding collision situation under different operation scenarios. Based on a pre-defined deep learning model, the fourth information is subjected to multi-objective optimization processing to obtain a target collision avoidance scheme. The simulation of the stacker-reclaimer's operation scenario is performed based on the second information, the third information, and the three-dimensional model to obtain the simulation results, including: Based on preset peripheral devices of different types, specifications and operating states, and corresponding preset operating rules, the second information is processed by device simulation to obtain a basic device simulation model. The third piece of information is analyzed and processed to obtain the equipment operation mode and potential collision risk points; By combining the basic simulation model of the equipment, the equipment operation mode, and potential collision risk points, predictive processing is performed to obtain abnormal simulation results under equipment failure and abnormal operation line conditions. Based on the three-dimensional model and the abnormal simulation results, optimization processing is performed. Combining the structural dimensions, operating speed and braking performance of the stacker-reclaimer and surrounding equipment, the response of the sensors and the operation adjustment strategies made by the stacker-reclaimer are predicted to obtain the simulation results of the operating scenario. Specifically, the simulation results of the operating scenarios are subjected to risk simulation processing to obtain fourth information. This fourth information includes the distribution of peripheral devices and corresponding collision situations under different operating scenarios, including: Based on the simulation results of the stacker-reclaimer's operation scenarios, a set of scenarios with different peripheral equipment and corresponding operating lines is obtained. The set of scenarios is identified based on a preset hidden Markov model to obtain the stacker-reclaimer operation mode identification result. Based on the operation mode recognition results of the stacker-reclaimer, a collaborative operation simulation is constructed. The stacker-reclaimer and peripheral equipment are treated as independent individuals, and their interaction behavior is simulated to obtain a collaborative operation simulation model. The operation collaborative simulation model is run and analyzed by a simulation engine, and the average number of braking times, collision prediction probability, and safe running time percentage are extracted to obtain the fourth information; The fourth information is subjected to multi-objective optimization processing based on a preset deep learning model to obtain a target collision avoidance scheme, including: Based on the fourth information and the preset Pareto front optimization model, modeling is performed to maximize safety, maximize operational efficiency and minimize energy consumption, resulting in at least two initial collision avoidance schemes. The performance of the initial collision avoidance scheme under various operating scenarios is simulated and evaluated based on a genetic algorithm to obtain the optimal collision avoidance scheme set. The optimal collision avoidance scheme set is processed based on a preset deep learning method. By setting a reward mechanism and a policy gradient method, the policy selection and dynamic adjustment are performed to obtain the target collision avoidance scheme.

2. The anti-collision method for the reclaimer arm of the stacker-reclaimer according to claim 1, characterized in that... Based on the first information, a 3D modeling process is performed to obtain a 3D model, including: Based on the first information, parametric modeling is performed to obtain an initial three-dimensional model; The initial three-dimensional model is meshed using adaptive mesh generation technology to obtain a finite element model. Modal analysis and static loading simulation analysis were performed on the finite element model to obtain the corresponding dynamic and static characteristics, respectively. Based on the dynamic characteristics, the static characteristics, the operating speed and braking performance of the stacker-reclaimer, the sensitivity data of the stacker-reclaimer are determined. Based on the finite element model, dynamic characteristics, static characteristics, and sensitivity data, a three-dimensional model of the stacker-reclaimer is constructed.

3. The anti-collision method for the reclaimer arm of the stacker-reclaimer according to claim 1, characterized in that... Based on a genetic algorithm, the performance of the initial collision avoidance scheme is simulated and evaluated in various scenarios to obtain the optimal collision avoidance scheme set, including: The initial population is initialized according to the initial anti-collision scheme to obtain an initial population, and the initial population is set under different operating scenarios. Each individual in the initial population represents an anti-collision scheme. The fitness of all individuals in the initial population in the same operating scenario is calculated to obtain the fitness of multiple individuals. The updated population is obtained by performing selection, crossover, and mutation operations based on the individual fitness values. The updated population is subjected to fitness evaluation and convergence test, and the final target population is obtained iteratively. The optimal individual and the corresponding optimal anti-collision scheme are determined from multiple final target populations corresponding to different operating scenarios, forming the set of optimal anti-collision schemes.

4. The anti-collision method for the reclaimer arm of the stacker-reclaimer according to claim 3, characterized in that... Based on a pre-defined deep learning approach, the optimal collision avoidance scheme set is processed. By setting a reward mechanism and a policy gradient method, policy selection and dynamic adjustment are performed to obtain the target collision avoidance scheme, including: Based on a preset deep learning model, feature extraction is performed on the set of optimal anti-collision schemes to obtain a set of feature vectors; The feature vector set is parsed to determine the key dimension information related to collision avoidance, and a mathematical model mapping the key dimension information and the preset reward standard is constructed. Initialize the policy space, and calculate the reward value of the candidate policies contained in the policy space based on the deep learning model and the mapping mathematical model; The gradient degree of each policy parameter is calculated based on the policy gradient algorithm, and the parameters of the candidate policy are adjusted according to the preset update rule to obtain the adjusted candidate policy. The adjusted candidate strategy reward value is calculated and convergence verification is performed. The optimal candidate strategy is obtained iteratively as the target collision avoidance scheme.

5. A collision avoidance system for the reclaimer arm of a stacker-reclaimer, characterized in that, include: The acquisition unit is used to acquire first information, second information and third information. The first information includes the structure and parameter information of the stacker-reclaimer. The second information includes the site layout, distribution of surrounding equipment and running trajectory information of the stacker-reclaimer's working area. The third information includes the historical operating data of the stacker-reclaimer. The first modeling unit is used to perform three-dimensional modeling processing based on the first information to obtain a three-dimensional model. The first simulation unit is used to simulate the operation scenario of the stacker-reclaimer based on the second information, the third information and the three-dimensional model, and obtain the operation scenario simulation results; The simulation unit is used to perform risk simulation processing on the simulation results of the operation scenario to obtain fourth information, which includes the distribution of peripheral devices and the corresponding collision situation under different operation scenarios. The first optimization unit is used to perform multi-objective optimization processing on the fourth information based on a preset deep learning model to obtain a target collision avoidance scheme. The first simulation unit includes: The second simulation unit is used to perform equipment simulation processing on the second information based on preset peripheral equipment of different types, specifications and operating states, as well as corresponding preset operating rules, to obtain a basic equipment simulation model. The first analysis unit is used to analyze and process the third information to obtain the equipment operation mode and potential collision risk points; The prediction processing unit is used to combine the equipment simulation basic model, the equipment operation mode and potential collision risk points to perform prediction processing and obtain abnormal simulation results under equipment failure and abnormal operation line conditions. The second optimization unit is used to perform optimization processing based on the three-dimensional model and the abnormal simulation results. It combines the structural dimensions, operating speed and braking performance of the stacker-reclaimer and surrounding equipment to predict the response of the sensors and the operation adjustment strategy made by the stacker-reclaimer, and obtain the simulation results of the operating scenario. Specifically, the simulation results of the operating scenarios are subjected to risk simulation processing to obtain fourth information. This fourth information includes the distribution of peripheral devices and corresponding collision situations under different operating scenarios, including: Based on the simulation results of the stacker-reclaimer's operation scenarios, a set of scenarios with different peripheral equipment and corresponding operating lines is obtained. The set of scenarios is identified based on a preset hidden Markov model to obtain the stacker-reclaimer operation mode identification result. Based on the operation mode recognition results of the stacker-reclaimer, a collaborative operation simulation is constructed. The stacker-reclaimer and peripheral equipment are treated as independent individuals, and their interaction behavior is simulated to obtain a collaborative operation simulation model. The operation collaborative simulation model is run and analyzed by a simulation engine, and the average number of braking times, collision prediction probability, and safe running time percentage are extracted to obtain the fourth information; The fourth information is subjected to multi-objective optimization processing based on a preset deep learning model to obtain a target collision avoidance scheme, including: Based on the fourth information and the preset Pareto front optimization model, modeling is performed to maximize safety, maximize operational efficiency and minimize energy consumption, resulting in at least two initial collision avoidance schemes. The performance of the initial collision avoidance scheme under various operating scenarios is simulated and evaluated based on a genetic algorithm to obtain the optimal collision avoidance scheme set. The optimal collision avoidance scheme set is processed based on a preset deep learning method. By setting a reward mechanism and a policy gradient method, the policy selection and dynamic adjustment are performed to obtain the target collision avoidance scheme.

6. The anti-collision system for the stacker-reclaimer arm according to claim 5, characterized in that, The first modeling unit includes: The second modeling unit is used to perform parametric modeling processing based on the first information to obtain an initial three-dimensional model; The meshing unit is used to perform meshing processing on the initial three-dimensional model based on adaptive meshing technology to obtain a finite element model; The second simulation unit is used to perform modal analysis and static loading simulation analysis on the finite element model to obtain the corresponding dynamic and static characteristics, respectively. The first determining unit is used to determine the sensitivity data of the stacker-reclaimer based on the dynamic characteristics, the static characteristics, the operating speed and braking performance of the stacker-reclaimer. The first construction unit is used to construct a three-dimensional model of the stacker-reclaimer based on the finite element model, dynamic characteristics, static characteristics, and sensitivity data.

Images