Crane identification system based on artificial intelligence

Abstract

The invention provides a crane identification system based on artificial intelligence, and belongs to the field of crane intellectualization. The artificial intelligence cargo identification subsystem, the intelligent hoisting path planning subsystem, the working environment sensing and adapting subsystem, the man-machine interaction subsystem and the displacement sensing subsystem are connected with the central control subsystem; the problems that an existing cargo recognition system is low in accuracy, cannot conduct real-time sensing and adaptation, and is low in overall performance and poor in operation safety are solved. Various technologies are creatively fused, automatic identification and intelligent planning are achieved, the operation efficiency and safety are improved, the cost is reduced, and the crane identification system can be suitable for multiple fields.

Description

Technical Field

[0001] The invention belongs to the field of crane intelligence, and in particular relates to a crane identification system based on artificial intelligence. Background Art

[0002] Traditional cranes have many limitations in cargo identification, lifting operations and environmental adaptation. Manual operation relies on experience, is prone to errors and safety hazards, and efficiency is difficult to further improve. With the rapid development of AI technology, its application in the crane field is expected to achieve more precise, efficient and safe operations. Summary of the invention

[0003] In view of the above-mentioned deficiencies in the prior art, the present invention provides an artificial intelligence-based crane identification system, which solves the problems of low accuracy, inability to perform real-time perception and adaptation, low overall performance and poor operational safety of the existing cargo identification system.

[0004] In order to achieve the above objectives, the technical solution adopted by the present invention is: an artificial intelligence-based crane identification system, comprising: a central control subsystem, and an artificial intelligence cargo identification subsystem connected to the central control subsystem, an intelligent lifting path planning subsystem, a working environment perception and adaptation subsystem, a human-computer interaction subsystem, and a displacement sensing subsystem;

[0005] The artificial intelligence cargo identification subsystem is used to identify the appearance characteristics, size and weight of the cargo using a deep learning algorithm and a high-resolution image sensor to obtain complete cargo information;

[0006] The intelligent lifting path planning subsystem is used to perform path simulation and planning based on cargo information using reinforcement learning to obtain the optimal lifting path;

[0007] The working environment perception and adaptation subsystem is used to measure the ambient temperature, humidity and wind speed information using preset sensors;

[0008] The human-machine interaction subsystem is used to utilize the human-machine interaction interface to exchange information between the operator and the crane identification system, and to manually intervene in the lifting operation to obtain human-machine interaction information;

[0009] The displacement sensing subsystem is used to monitor the displacement information of the moving parts in real time by installing the displacement sensors on the moving parts of the crane;

[0010] The central control subsystem is used to receive and process data from each subsystem, and issue control instructions to the crane's actuator according to the processed data.

[0011] The beneficial effects of the present invention are as follows: the present invention solves the problems of low accuracy, inability to perceive and adapt in real time, low overall performance and poor operational safety of existing cargo identification systems by utilizing deep learning algorithms of artificial intelligence technology and convolutional neural networks in conjunction with a variety of sensors; and innovatively integrates a variety of technologies to achieve the advantages of automatic identification and intelligent planning, improve operational efficiency and safety, and reduce costs, so that the crane identification system of the present invention can be applied to multiple fields.

[0012] Furthermore, the central control subsystem receives and processes data from each subsystem specifically as follows:

[0013] Receive cargo information from the artificial intelligence cargo identification subsystem and coordinate the lifting process based on the cargo information;

[0014] Receive the path instruction of the intelligent lifting path planning subsystem and convert the path instruction into a control signal for the actuator;

[0015] Receive environmental data from the working environment perception and adaptation subsystem and adjust the crane operating parameters according to preset thresholds;

[0016] Receive the optimal lifting path information planned by the intelligent lifting path planning subsystem and generate lifting instructions;

[0017] Receive displacement information fed back by the displacement sensor system and control the lifting operation;

[0018] The central control subsystem issues control instructions to the crane's actuator based on cargo information, actuator control signals, crane operating parameters, lifting instructions, and displacement information, driving the crane to move along the planned path.

[0019] The beneficial effects of the above further scheme are as follows: the present invention utilizes the communication interface to communicate data with each subsystem, receives, processes and analyzes the data of each subsystem, issues control instructions to the crane actuator according to the processing results, coordinates the operation of each subsystem, and realizes the intelligent lifting operation of the crane identification system; the integrated artificial intelligence algorithm engine further improves the accuracy of decision-making; the present invention provides a central control subsystem that centrally controls and coordinates various functions; the central control subsystem of the present invention improves the closeness of the collaborative work of each subsystem and the processing capability of intelligent decision-making, and improves the ability of the crane identification system to integrate the functions of each subsystem.

[0020] Furthermore, the artificial intelligence cargo identification subsystem includes:

[0021] An image acquisition module is used to obtain image information of the cargo by placing a high-definition camera around the crane hook;

[0022] An image recognition module is used to process the acquired cargo image information using a convolutional neural network to identify the type, shape, and size of the cargo;

[0023] A weight detection module is used to measure the weight of the cargo by connecting the second sensor to the crane hook;

[0024] The data integration module is used to integrate the type, shape, size information and weight of the object to generate complete cargo characteristic information.

[0025] The beneficial effects of the above further scheme are as follows: the present invention realizes accurate identification of cargo type, shape and size information by processing image information obtained by the first sensor using a convolutional neural network algorithm; and generates complete cargo feature information in combination with the weight information measured by the second sensor, thereby improving the accuracy and comprehensiveness of cargo identification, providing a more accurate basis for subsequent lifting operations, and helping to improve the safety and efficiency of lifting operations; compared with the traditional manual method of identifying cargo, the present invention greatly reduces human errors and improves the identification speed; the cargo identification of the present invention not only relies on visual information, but also combines weight information, which can grasp the characteristics of cargo more comprehensively, and the neural network algorithm is more intelligent and accurate in identifying cargo types.

[0026] Furthermore, the intelligent lifting path planning subsystem includes: an environment detection module and a path planning and optimization module;

[0027] The environment detection module is used to obtain the environment information around the crane using laser radar and ultrasonic sensors;

[0028] The path planning and optimization module is based on a preset artificial intelligence algorithm and is used to combine three-dimensional environmental information, close-range obstacle information, complete cargo feature information and the performance parameters of the crane itself. During the lifting process, the lifting path is adjusted in real time according to environmental changes to generate the optimal lifting path.

[0029] Furthermore, the environment detection module specifically includes:

[0030] The radar scanning submodule is used to scan the working space and obtain three-dimensional environmental information by installing the laser radar on the road working area of ​​the crane;

[0031] The ultrasonic detection submodule uses ultrasonic waves to obtain obstacle information at close range by installing ultrasonic sensors on the hook, boom or body of the crane.

[0032] The beneficial effects of the above further scheme are as follows: the present invention realizes the planning of the optimal path for lifting cargo by combining environmental information, cargo feature information and the crane's own performance parameters obtained by laser radar and ultrasonic sensors based on reinforcement learning or other suitable preset artificial intelligence algorithms, and can adjust the path in real time according to environmental changes during the lifting process; improves the rationality and safety of the lifting path, effectively avoids the occurrence of accidents such as collisions, and improves the lifting efficiency; compared with the traditional method of planning the lifting path based on manual experience, the intelligent lifting path planning subsystem of the present invention can cope with more complex work scenarios and dynamic changes; the present invention considers more comprehensive factors when planning the path, the algorithm is more complex and the path can be adjusted dynamically in real time, and the comprehensive utilization of environmental information is higher.

[0033] Furthermore, the execution steps of the preset artificial intelligence algorithm adopted by the path planning and optimization module are as follows:

[0034] Based on 3D environmental information, close-range obstacle information, complete cargo feature information, and the crane’s own performance parameters;

[0035] Initialize the reinforcement learning algorithm, set the state space of the reinforcement learning algorithm, and use the three-dimensional environment information, close-range obstacle information, complete cargo feature information, and the performance parameters of the crane itself as variables in the state space;

[0036] Define the action space and speed adjustment range, and set the reward function based on the lifting path length, collision risk, and lifting efficiency;

[0037] In the initial state, an action is selected from the action space according to the current crane state information, and the selected action is executed. The reward value is calculated according to the changed environment state and the set reward function;

[0038] Adjust the strategy based on reward feedback, conduct experiments and learn in a simulated environment, and after multiple iterations, learn the optimal strategy in various situations;

[0039] The lifting path is adjusted in real time based on the learned optimal strategy, the current crane parameters, and the dynamic changes in the environment to generate the optimal lifting path from the initial location of the cargo to the target location.

[0040] The beneficial effects of the above further scheme are as follows: the present invention solves the problems of low accuracy, inability to perceive and adapt in real time, low overall performance and poor operational safety of existing cargo identification systems by utilizing deep learning algorithms of artificial intelligence technology and convolutional neural networks in conjunction with a variety of sensors; and innovatively integrates a variety of technologies to achieve the advantages of automatic identification and intelligent planning, improve operational efficiency and safety, and reduce costs, so that the crane identification system of the present invention can be applied to multiple fields.

[0041] Further, the working environment perception and adaptation subsystem includes: an environment monitoring module and a parameter adjustment module;

[0042] The environment monitoring module is used to detect environmental information of the working environment using an environmental sensor;

[0043] The parameter adjustment module is used to adjust the operating parameters of the crane according to the ambient temperature, humidity and wind speed information and preset thresholds of the corresponding environmental information.

[0044] Furthermore, the environmental monitoring module specifically includes:

[0045] The temperature monitoring submodule is used to monitor the temperature information of the working environment in real time by installing temperature sensors around the motor and electrical control cabinet;

[0046] The humidity monitoring submodule is used to monitor the humidity information of the working environment in real time by installing humidity sensors in the concentrated area of ​​electrical equipment;

[0047] The wind speed monitoring submodule is used to monitor the wind speed information of the working environment in real time by installing a wind speed sensor on the top of the crane.

[0048] The beneficial effects of the above further scheme are as follows: the present invention uses a variety of sensors to conduct all-round monitoring of the working environment, and uses artificial intelligence technology to analyze and process the sensor data, thereby realizing real-time identification of obstacles, prediction of environmental change trends, and automatic adjustment of crane operating parameters and operation strategies according to environmental perception results; improving the adaptability of the crane to different working environments, ensuring the smooth progress of operations under various environmental conditions, and reducing safety risks caused by environmental factors; traditional cranes have limited environmental perception capabilities and it is difficult to achieve such comprehensive environmental monitoring and adaptive adjustment; the present invention is more diverse in environmental information collection, and the adaptive adjustment to environmental changes is smarter and more comprehensive. It is not only limited to obstacle avoidance functions, but can also adjust operating parameters according to environmental factors.

[0049] Furthermore, the human-computer interaction subsystem includes:

[0050] The information display module is used to display cargo information, lifting path, working environment parameters and system operation status information by providing a human-computer interaction interface;

[0051] The operation input module provides operation buttons and a touch screen to input manual lifting task instructions, manually intervene in lifting operations, and feed back human-computer interaction information to the central control subsystem.

[0052] The beneficial effects of the above further scheme are as follows: the present invention provides a human-computer interaction interface, operation buttons and a touch screen, which are convenient for operators to monitor in real time and intervene manually; the overall coordination and intelligence of the crane identification system are improved, ensuring that the subsystems work closely together to efficiently complete the lifting task; the display unit displays cargo information, lifting path, working environment parameters and system operation status information in real time, and the operation unit allows the operator to intervene manually, thereby realizing information interaction between the operator and the system; and the operator's ability to control the lifting operation is improved, which is convenient for the operator to understand the operation status in time and make adjustments when necessary, thereby enhancing the flexibility and reliability of the system; the present invention provides comprehensive information display and operation intervention means; the human-computer interaction interface of the present invention has more advantages in the integrity of information display and the convenience of operation, and can better meet the needs of operators. BRIEF DESCRIPTION OF THE DRAWINGS

[0053] Figure 1 The figure is a flow chart of the method of the present invention.

[0054] Figure 2 Schematic diagram of the position and working area of ​​each sensor in this embodiment.

[0055] Figure 3 This is a flowchart of the work flow in this embodiment.

[0056] Figure 4 This is the interaction between sensors in this embodiment.

[0057] Figure 5 This is a diagram of the types of sensors used in this embodiment.

[0058] Figure 6 2 is a network architecture diagram of the convolutional neural network in this embodiment. DETAILED DESCRIPTION

[0059] The specific implementation modes of the present invention are described below so that those skilled in the art can understand the present invention. However, it should be clear that the present invention is not limited to the scope of the specific implementation modes. For those of ordinary skill in the art, as long as various changes are within the spirit and scope of the present invention as defined and determined by the attached claims, these changes are obvious, and all inventions and creations utilizing the concept of the present invention are protected.

[0060] Before describing the present invention, the following terms are explained:

[0061] CNN: Convolutional Neural Network;

[0062] AI: Artificial Intelligence.

[0063] Example 1

[0064] like Figure 1 As shown, the present invention provides an artificial intelligence-based crane identification system, including: a central control subsystem, and an artificial intelligence cargo identification subsystem connected to the central control subsystem, an intelligent lifting path planning subsystem, a working environment perception and adaptation subsystem, a human-computer interaction subsystem and a displacement sensing subsystem.

[0065] In this embodiment, an artificial intelligence-based crane identification system is provided, comprising: a central control subsystem, and an artificial intelligence cargo identification subsystem, an intelligent lifting path planning subsystem, a working environment perception and adaptation subsystem, a human-computer interaction subsystem, and a displacement sensing subsystem connected to the central control subsystem;

[0066] Artificial intelligence cargo identification subsystem: uses deep learning algorithms and high-resolution image sensors to accurately identify the appearance, size, weight, etc. of cargo;

[0067] Intelligent lifting path planning subsystem: Based on the AI-based path planning algorithm, it comprehensively considers the mechanical performance of the crane, the workspace layout, the characteristics of the cargo, and the real-time environmental information to plan the optimal lifting path;

[0068] Work environment perception and adaptation subsystem: Uses a variety of sensors (such as lidar, ultrasonic sensors, temperature and humidity sensors, etc.) to monitor the work environment in all directions; measures ambient temperature, humidity, and wind speed information;

[0069] Human-computer interaction subsystem: realizes information interaction between operators and the system, including displaying lifting-related information and receiving manual operation instructions from operators, and enables operators to manually intervene in lifting operations and obtain human-computer interaction information;

[0070] Displacement sensing subsystem: By installing the displacement sensor on the moving parts of the crane, when receiving the lifting task instruction, during the lifting process, the displacement sensor is used to monitor the displacement information of key moving parts such as the hook in real time and feed it back to the central control system to provide operation feedback to the operator.

[0071] In this embodiment, by integrating multiple advanced technologies, accurate identification of cargo, intelligent lifting path planning, and real-time perception and adaptation to the working environment are achieved, thereby significantly improving the overall performance and operational safety of the crane.

[0072] The central control subsystem receives and processes the data of each subsystem specifically as follows:

[0073] Receive cargo information from the artificial intelligence cargo identification subsystem and coordinate the lifting process based on the cargo information;

[0074] Receive the path instruction of the intelligent lifting path planning subsystem and convert the path instruction into a control signal for the actuator;

[0075] Receive environmental data from the working environment perception and adaptation subsystem and adjust the crane operating parameters according to preset thresholds;

[0076] Receive the optimal lifting path information planned by the intelligent lifting path planning subsystem and generate lifting instructions;

[0077] Receive displacement information fed back by the displacement sensor system and control the lifting operation;

[0078] The central control subsystem issues control instructions to the crane's actuators based on cargo information, actuator control signals, crane operating parameters, lifting instructions, and displacement information, driving the crane's components to move along the planned path.

[0079] In this embodiment, the central control subsystem includes:

[0080] Data communication module: Use the communication interface to carry out stable and efficient data transmission with other subsystems to ensure timely interaction of information;

[0081] Data processing and analysis module: Receives data from each subsystem, uses advanced algorithms and models to perform in-depth processing and comprehensive analysis, and extracts key information for decision-making;

[0082] Control instruction generation module: Generates accurate control instructions for the crane actuator based on the results of processing and analysis, including the setting of parameters such as movement direction and speed;

[0083] Coordination management module: coordinates the operation sequence and work rhythm among subsystems such as the artificial intelligence cargo identification subsystem, the intelligent lifting path planning subsystem, and the working environment perception and adaptation subsystem to ensure the overall efficient collaboration of the system;

[0084] The central control subsystem uses the communication interface to receive cargo information from the artificial intelligence cargo identification subsystem and coordinate the lifting process accordingly;

[0085] Receive the path instructions of the intelligent lifting path planning subsystem and convert them into control signals for the actuators;

[0086] Obtain environmental data from the working environment perception and adaptation subsystem and adjust crane operating parameters according to thresholds;

[0087] Receive the optimal lifting path information planned by the intelligent lifting path planning subsystem, including detailed instructions such as the movement trajectory of the hook (such as coordinates, speed, and acceleration changes over time), lifting height, rotation angle, etc.;

[0088] According to the displacement information fed back by the displacement sensor, the lifting operation is controlled, including: adjusting the movement direction and speed of the components, and using the displacement information to issue collision warnings and prevent the crane from tipping over;

[0089] Based on this information, the central control subsystem will issue precise control instructions to the crane's actuators (such as motors, hydraulic systems, etc.), driving the crane components to move along the planned path to ensure the accuracy and safety of the lifting operation.

[0090] The artificial intelligence cargo identification subsystem includes:

[0091] An image acquisition module is used to obtain image information of the cargo by placing a high-definition camera around the crane hook;

[0092] An image recognition module is used to process the acquired cargo image information using a convolutional neural network to identify the type, shape, and size of the cargo;

[0093] The weight detection module is used to measure the weight of the cargo by connecting a high-precision weighing sensor to the crane hook;

[0094] The data integration module is used to integrate the type, shape, size information and weight of the object to generate complete cargo characteristic information.

[0095] In this embodiment, the artificial intelligence cargo identification subsystem is driven by AI and includes: an image acquisition module, which uses at least one high-definition image sensor and is arranged around the crane hook to obtain image information of the cargo;

[0096] The image recognition module is used to process the acquired image information using a convolutional neural network to identify the type, shape and size of the goods, and continuously improve the recognition accuracy through training with a large amount of image data;

[0097] The weight detection module measures the weight of the cargo by connecting a high-precision weighing sensor to the crane hook to ensure the accuracy of the weight data;

[0098] The data integration module is used to combine the weight information obtained by the high-precision weighing sensor with the type, shape and size information of the goods identified by the convolutional neural network algorithm module to generate complete cargo feature information;

[0099] The artificial intelligence cargo identification subsystem uses a convolutional neural network (CNN) to train a large number of cargo images, enabling it to automatically extract the key features of the cargo and accurately determine the type of cargo. It is equipped with a high-precision weighing sensor to combine weight data with image recognition results to provide comprehensive and accurate cargo information for subsequent lifting operations.

[0100] The intelligent lifting path planning subsystem includes: an environment detection module and a path planning and optimization module;

[0101] The environment detection module is used to obtain the environment information around the crane, specifically including:

[0102] The radar scanning submodule is used to scan the working space and obtain three-dimensional environmental information by installing the laser radar on the road working area of ​​the crane;

[0103] Ultrasonic detection submodule, which uses ultrasonic waves to obtain obstacle information at close range by installing ultrasonic sensors on the hook, boom or body of the crane;

[0104] The path planning and optimization module is based on a preset artificial intelligence algorithm and is used to combine three-dimensional environmental information, close-range obstacle information, complete cargo feature information and the performance parameters of the crane itself. During the lifting process, the lifting path is adjusted in real time according to environmental changes to generate the optimal lifting path.

[0105] In this embodiment, the intelligent lifting path planning subsystem includes: an environment detection module and a path planning and optimization module;

[0106] The environment detection module comprises:

[0107] The laser radar scanning submodule is used to scan the working space and obtain three-dimensional environmental information, including the location, shape and distance data of obstacles by installing the laser radar in the road working area of ​​the crane;

[0108] Ultrasonic detection submodule, by installing ultrasonic sensors on the hook, boom or body of the crane, is used to detect obstacles at close range, making up for the shortcomings of laser radar in close range detection and ensuring the integrity of environmental information;

[0109] AI path planning algorithm module, based on reinforcement learning or preset AI algorithm;

[0110] The preset AI algorithm can select a fuzzy logic algorithm when there is uncertainty or ambiguity in the information in the working environment, such as when the measurement of environmental parameters is not accurate enough, or when there is a certain fuzzy range in the judgment of the shape and position of obstacles; when in a large storage area, cranes need to frequently lift goods between many shelves and aisles, and the environment may change dynamically with the stacking and handling of goods, an ant colony algorithm can be used; it has strong adaptability and distributed computing capabilities, and can gradually find the optimal path from the source point of the goods to the target point through information exchange and collaboration between individual ants, and can quickly readjust the path when the environment changes, effectively avoid obstacles, and improve lifting efficiency;

[0111] The AI ​​algorithm takes into account many factors during the training process, such as the crane's dynamic model, cargo characteristics, environmental obstacle distribution, and lifting efficiency. It combines the environmental information obtained by the lidar and ultrasonic sensors, the complete cargo feature information generated by the artificial intelligence cargo identification subsystem, and the crane's own performance parameters to quickly and accurately plan the optimal lifting path from the initial location of the cargo to the target location, and can adjust the path in real time according to environmental changes during the lifting process.

[0112] The specific steps of executing the preset artificial intelligence algorithm are as follows: collecting data information, using laser radar to scan the workspace to obtain three-dimensional environmental information, including the precise position of obstacles (expressed in three-dimensional coordinates), shape (identified by the characteristics of reflected light) and distance data (calculated based on the flight time of the laser); obtaining close-range obstacle information through ultrasonic sensors to supplement the blind spots that may exist in the laser radar; receiving complete cargo feature information from the artificial intelligence cargo identification subsystem, wherein cargo images are collected through high-resolution image sensors set around the crane hook, and the type of cargo (compared with a large amount of pre-trained cargo image data), shape (extracting edge and contour features) and size information (calculated based on image ratio and reference size) are identified by image recognition module processing, and the cargo weight is measured by high-precision weighing sensors, and the above information is integrated by the data integration module; and the performance parameters of the crane itself are obtained, such as the maximum lifting weight, maximum lifting height, maximum amplitude, lifting and operating speed range;

[0113] Initialize the reinforcement learning algorithm, set the state space of the reinforcement learning algorithm, and use the above-collected three-dimensional environmental information, close-range obstacle information, complete cargo feature information, and the performance parameters of the crane itself as variables in the state space; define the action space, such as the moving direction of the hook (up and down, left and right, front and back, etc.), speed adjustment range, etc.; determine the reward function, comprehensively considering factors such as the length of the lifting path (the shorter the path, the higher the reward), collision risk (the closer the distance to the obstacle, the greater the penalty, and the higher the reward for no collision), and lifting efficiency (the shorter the lifting time, the higher the reward);

[0114] The deep learning algorithm is trained in a loop. In the initial state, the agent of the deep learning algorithm selects an action from the action space based on the current state information, for example, determining the initial moving direction and speed of the hook. After executing the action, the state of the environment changes, and the system calculates the reward value based on the new state and the pre-set reward function. The agent adjusts its strategy based on the reward feedback, and through continuous experiments and learning in the simulation environment, after multiple iterations, it gradually learns the optimal strategy in various situations, that is, generates the optimal lifting path from the initial position of the cargo to the target position.

[0115] During the lifting process, as the environment changes dynamically (such as the movement of obstacles, changes in wind speed, etc.), the sensors update data information in real time, and the intelligent agent can adjust its actions in time according to the new status and optimize the path in real time to ensure that the lifting operation always remains efficient and safe;

[0116] The intelligent lifting path planning subsystem uses a reinforcement learning algorithm to enable the crane identification system to learn and optimize the path planning strategy in continuous simulation operations to adapt to different working scenarios and dynamic changes. It also interacts with the crane's central control subsystem in real time to dynamically adjust the path according to actual conditions, ensuring the efficiency and safety of the lifting process and obtaining the optimal lifting path.

[0117] The working environment perception and adaptation subsystem includes: an environment monitoring module and a parameter adjustment module;

[0118] The environment monitoring module is used to detect the environmental information of the working environment, specifically including:

[0119] The temperature monitoring submodule is used to monitor the temperature information of the working environment in real time by installing temperature sensors around the motor and electrical control cabinet;

[0120] The humidity monitoring submodule is used to monitor the humidity information of the working environment in real time by installing humidity sensors in the concentrated area of ​​electrical equipment;

[0121] The wind speed monitoring submodule is used to monitor the wind speed information of the working environment in real time by installing a wind speed sensor on the top of the crane;

[0122] The parameter adjustment module is used to adjust the operating parameters of the crane according to the ambient temperature, humidity and wind speed information and preset thresholds of the corresponding environmental information.

[0123] In this embodiment, the working environment perception and adaptation subsystem includes: an environment monitoring module and a parameter adjustment module;

[0124] The environment monitoring module includes: a temperature monitoring submodule, a humidity monitoring submodule and a wind speed monitoring submodule, which respectively monitor the temperature, humidity and wind speed information of the working environment in real time by installing a temperature sensor around the motor and the electrical control cabinet, a humidity sensor in the concentrated area of ​​the electrical equipment, and a wind speed sensor on the top of the crane;

[0125] A parameter adjustment module, used to adjust the operating parameters of the crane according to the temperature, humidity and wind speed information of the working environment and the preset threshold values, which are shown in Table 1, including reducing the operating speed or taking other safety measures in a harsh environment;

[0126] The working environment perception and adaptation subsystem uses AI technology to analyze and process sensor data, identify obstacles in real time, and predict environmental change trends (such as wind speed changes, light intensity changes, etc.);

[0127] Automatically adjust the operating parameters and operation strategies of the crane according to the environmental perception results to ensure smooth operation under various environmental conditions;

[0128] Table 1

[0129]

[0130] The human-computer interaction subsystem includes:

[0131] The information display module provides a display screen for displaying cargo information, lifting path, working environment parameters and system operation status information;

[0132] The operation input module provides operation buttons and a touch screen to input manual lifting task instructions, manually intervene in lifting operations, and feed back human-computer interaction information to the central control subsystem.

[0133] In this embodiment, the human-computer interaction interface includes:

[0134] The information display module provides a display unit, such as a display screen, for clearly displaying cargo information, lifting path, working environment parameters and system operation status information in real time, so that operators can understand the details of the lifting operation at any time;

[0135] The operation input module provides operation units such as operation buttons and touch screens. Operators can use this module to manually input lifting task instructions (including the starting position and target position of cargo lifting, special requirements such as lifting speed limit and cargo placement posture, etc.), as well as manually intervene in the lifting operation (such as pausing lifting, adjusting the path, emergency braking, etc.), and promptly feed back human-computer interaction information to the central control subsystem.

[0136] In this embodiment, after the crane identification system receives the lifting task instruction, during the lifting process, the displacement sensor monitors the displacement information of key moving parts such as the hook in real time and feeds it back to the central control system to provide operation feedback for the operator; when the hook moves to a position roughly above the cargo, the high-definition image sensor captures the cargo image and transmits it to the artificial intelligence cargo identification subsystem, and its CNN algorithm processes the image and identifies the cargo information in combination with the cargo database. At the same time, the high-precision weighing sensor measures the cargo weight, and the information of the two is integrated and transmitted to the central control system; when planning the lifting path, the laser radar scans the workspace to obtain three-dimensional environmental information and transmits it to the lifting path planning subsystem. The ultrasonic sensor detects close-range obstacle information and supplements blind spot detection data, which is also transmitted to the module. The temperature, humidity and wind speed sensors monitor environmental information and transmit it to the working environment perception and adaptation subsystem, and then share it with the intelligent lifting path planning subsystem. The AI ​​path planning algorithm combines various information to calculate the optimal lifting path. The path planning result is displayed on the human-computer interaction interface. The operator can confirm or adjust the path. After that, the central control subsystem issues control instructions to the crane actuator based on the path and real-time environmental information. During the lifting operation, each sensor continuously monitors. Once the environmental changes affect the lifting operation, the central control subsystem will adjust the lifting path in time.

[0137] The present invention integrates an AI algorithm engine to perform in-depth analysis and decision processing on the data from each module, and issue accurate control instructions to the crane's actuator;

[0138] The present invention is provided with a human-computer interaction interface, through which an operator can monitor the system operation status in real time, view data analysis results, and perform necessary manual intervention.

[0139] Example 2

[0140] In this embodiment, Figure 2 As shown in the figure, the positions of the sensors and the working area diagram of the present invention are shown, and the crane identification system is powered on and self-checked, such as Figure 3 As shown in the figure, start the crane power supply and the system starts to be powered on; each subsystem and sensor immediately starts the self-test program as shown in the figure. Figure 4As shown in the figure, the working process of each sensor is as follows: the high-definition image sensor checks whether the lens, image acquisition circuit and communication module are working properly to ensure that the image data can be collected and transmitted normally; the laser radar checks the internal circuit and rotating scanning mechanism to verify that its transmission and receiving functions are normal and the scanning angle and range meet the set values; the ultrasonic sensor detects the signal transmission and receiving functions to ensure the accuracy of close-range obstacle detection; the temperature, humidity and wind speed sensors calibrate the zero point and range, and check the data transmission interface; the displacement sensor confirms the installation position and measurement accuracy, and checks the connection with the central control system; the high-precision weighing sensor calibrates the zero point and linearity, and checks the connection stability and signal transmission between the sensor and the hook.

[0141] In this embodiment, Figure 5 As shown, the sensors used for environmental perception in the present invention include: a humidity sensor, a wind speed sensor, a radar, a high-definition camera, a weight sensor, a temperature sensor, and a displacement sensor.

[0142] In this embodiment, the communication link test and initialization are performed: a communication link is established between the central control subsystem and each module and sensor, and the stability of the link and the accuracy of data transmission are checked by sending and receiving test signals; technologies such as CRC (cyclic redundancy check) are used to ensure that no errors occur in the data transmission process; at the same time, the communication protocol is initialized, and parameters such as the format, rate and priority of data transmission are set;

[0143] Loading the AI ​​algorithm module: In the artificial intelligence cargo identification subsystem, the image recognition module uses the convolutional neural network (CNN) algorithm to load the pre-trained model parameters, including the weights and biases of the convolution kernels, to ensure that the model can accurately identify the characteristics of the cargo; the AI ​​path planning algorithm (such as the reinforcement learning algorithm) of the lifting path planning module loads parameters such as the learning rate, reward function, state space, and action space, and prepares for path planning calculations; at the same time, the crane identification system allocates corresponding space in the memory for data processing and storage, preparing for subsequent operations; after the crane identification system completes initialization, it enters the standby state, and the human-computer interaction interface displays the system ready information, waiting for lifting task instructions.

[0144] When starting a transportation task, the command is first received: the operator manually inputs the lifting task command through the human-computer interaction interface, or the system receives the lifting task command from the superior dispatching system (such as the logistics management system, port operation management system, etc.) through the network communication interface; the command content includes the starting position of the cargo lifting (such as the coordinates of the cargo in the warehouse, the shelf position, etc.) and the target position (such as the specific location of the unloading area, the loading point on the transport vehicle, etc.), and may also include some special requirements, such as lifting speed limit, cargo placement posture, etc.;

[0145] Prepare for hook movement: The central control subsystem preliminarily plans the approximate direction and distance of the hook movement according to the lifting task instructions, and sends control instructions to the crane's actuators (motors, hydraulic systems, etc.). The actuators drive the crane's trolley, car, hook and other components to start moving, so that the hook moves closer to the direction of the cargo; during this process, the displacement sensor monitors the displacement data of key components such as the hook and boom in real time, and feeds the data back to the central control subsystem; the central control subsystem adjusts the action of the actuator according to the displacement feedback information to ensure that the hook can move accurately in the direction of the cargo while avoiding collision with the surrounding environment.

[0146] In cargo identification, image acquisition and preprocessing are performed: when the hook approaches the cargo, the high-definition image sensor in the image acquisition module begins to acquire images of the cargo; during the acquisition process, the high-definition image sensor automatically adjusts parameters such as exposure time and gain according to light conditions to obtain clear images; the acquired image data is transmitted to the artificial intelligence cargo identification subsystem through a communication link; in the artificial intelligence cargo identification subsystem, the image is first preprocessed, including image denoising, grayscale (if grayscale image recognition is used) and image enhancement operations to improve image quality and facilitate subsequent feature extraction;

[0147] The preprocessed image is then recognized by the CNN algorithm module, such as Figure 6 As shown in the figure, the preprocessed image is input into the convolutional neural network (CNN) algorithm. The convolution layer performs convolution operations with the image through multiple different convolution kernels to extract local features in the image, such as edges and textures. The pooling layer is used to downsample the convolutional features to reduce the amount of data while retaining key features. After being processed by multiple convolution and pooling layers, the fully connected layer is used to combine the extracted features to form a high-level feature representation of the goods. Finally, the output layer compares the high-level feature representation with the features in the pre-trained goods database to identify the type, shape and size information of the goods.

[0148] Then weight measurement and information integration are carried out: in response to the pre-processed image being recognized by the CNN algorithm module of the image recognition module, the high-precision weighing sensor of the weight detection module measures the weight of the goods in real time; the strain gauge inside the weighing sensor converts the pressure change into an electrical signal through circuits such as a Wheatstone bridge according to the pressure change caused by the gravity of the goods, and after amplification, filtering and analog-to-digital conversion, accurate weight data is obtained and transmitted to the goods identification module; the data integration module integrates the cargo image information recognized by the CNN algorithm module with the weight information measured by the weighing sensor to generate complete cargo feature information, including cargo name, shape, size and weight, etc., and transmits the complete cargo feature information to the central control subsystem.

[0149] The transportation path planning is firstly carried out through the laser radar scanning of the radar scanning submodule: the laser radar performs an all-round scan of the entire workspace at a certain scanning frequency; the laser beam emitted by the laser radar is reflected back after encountering an obstacle, and the distance between the obstacle and the laser radar is calculated based on the flight time of the laser; through continuous scanning, the laser radar obtains the three-dimensional coordinate information, shape and distance data of all obstacles in the workspace, and constructs a three-dimensional environmental model of the workspace; the three-dimensional environmental information is transmitted to the intelligent lifting path planning subsystem in real time through a high-speed communication link;

[0150] Ultrasonic sensor detection: ultrasonic sensors installed on the crane hook, boom or body work simultaneously, emitting ultrasonic signals to the surroundings at a higher frequency; when the ultrasonic wave encounters a close obstacle, it will be reflected back, and the sensor calculates the distance to the obstacle based on the round-trip time of the ultrasonic wave; the detection data of the ultrasonic sensor supplements the blind area and close-range area scanned by the lidar, further improving the environmental information of the workspace, especially focusing on the areas prone to obstacles such as under the hook and around the crane boom, ensuring the integrity of the environmental information and obtaining close-range obstacle information; the close-range obstacle information is also transmitted to the intelligent lifting path planning subsystem in real time;

[0151] Path planning calculation: In the intelligent lifting path planning subsystem, the AI ​​path planning algorithm (such as reinforcement learning algorithm) of the path planning and optimization module calculates the optimal lifting path from the current position to the target lifting position based on the cargo information (type, shape, size, weight) transmitted by the central control subsystem, the current position and posture information of the crane (the position of the trolley and the trolley, the angle and length of the boom, etc.), combined with the environmental information (the position, shape and distance of obstacles) collected by the laser radar and ultrasonic sensors, and comprehensively considers the performance parameters of the crane (maximum lifting weight, maximum lifting height, maximum amplitude, lifting and operating speed range, etc.), the characteristics of the cargo (center of gravity position, stability requirements, etc.), the distribution of obstacles in the environment, and the influence of environmental parameters on the lifting operation (such as the influence of wind speed on lifting stability) and other complex factors. During the path planning process, the algorithm searches for the optimal path by continuously evaluating the "reward value" of different paths (such as a comprehensive evaluation of factors such as the shortest path length, the lowest collision risk, and the highest lifting efficiency). The calculated path information includes the hook's moving trajectory (coordinates, speed, acceleration changes over time), lifting height, rotation angle and other detailed instructions. These instructions are transmitted to the central control system in the form of data and displayed in a visual way on the human-computer interaction interface for the operator to view. Finally, the path is confirmed and adjusted, and the path is adjusted in real time according to environmental changes during the lifting process.

[0152] Lifting execution, operation command generation and sending: After the operator confirms the path on the human-machine interface, the central control subsystem issues detailed control instructions to the crane's actuators (motor, hydraulic system, etc.) based on the planned optimal lifting path and real-time environmental information; these instructions include the running speed and direction of the trolley and carriage, the telescopic length and angle change speed of the boom, and the lifting and lowering speed of the hook, etc. The instructions are transmitted to the corresponding actuator controller through the communication link between the control system and the actuator (such as CAN bus, industrial Ethernet, etc.);

[0153] Component motion detection: The crane's actuator drives the boom, hook and other components to move along a predetermined path according to control instructions; in the displacement sensor subsystem, the displacement sensor monitors the displacement information of these components in real time, for example: the linear displacement sensor measures the change in the lifting height of the hook, and the angle displacement sensor monitors the change in the rotation angle of the boom; the displacement sensor feeds back the measured displacement data to the central control subsystem through a high-speed communication link; the control algorithm in the central control subsystem accurately adjusts the movement speed and direction of the actuator according to the deviation between the displacement feedback data and the predetermined path through a PID (proportional-integral-differential) controller or other advanced control algorithms to ensure the accuracy and stability of the lifting operation; for example, if the actual lifting height of the hook deviates from the predetermined height, the control system will adjust the speed of the hook lifting motor to return it to the predetermined height;

[0154] Environmental perception and path adjustment: During the lifting process, LiDAR and ultrasonic sensors continuously scan the working environment and detect changes in the position of obstacles in real time;

[0155] Environmental parameter information is continuously collected through temperature, humidity, and wind speed sensors; temperature sensors use components such as thermistors to measure ambient temperature, humidity sensors obtain humidity values ​​by measuring changes in capacitance or resistance with humidity, and wind speed sensors measure wind speed based on the force of wind on the wind cup or hot wire inside it. The collected temperature, humidity, and wind speed data are transmitted to the working environment perception and adaptation subsystem, which analyzes and processes the data to determine whether the current environment has an impact on the crane's lifting operations; if environmental parameters exceed the preset safety threshold, the central control subsystem will take corresponding measures (such as reducing the lifting speed, issuing environmental alarms, etc.), and share environmental parameter information with the intelligent lifting path planning subsystem for real-time adjustments;

[0156] If the lidar or ultrasonic sensor detects the appearance of a new obstacle or the position of the original obstacle changes, the intelligent lifting path planning subsystem will recalculate the lifting path based on the new environmental information; at the same time, the working environment perception and adaptation subsystem determines whether the operating parameters of the crane need to be adjusted according to the changes in environmental parameters; for example, if the wind speed suddenly increases and exceeds the safety threshold, the working environment perception and adaptation subsystem will notify the central control subsystem to reduce the lifting speed or adjust the boom angle to enhance stability; the central control subsystem will adjust the operation of the crane actuator in a timely manner according to the new path planning results and environmental parameter adjustment instructions to ensure the smooth completion of the lifting operation; during the lifting process, the human-computer interaction subsystem displays the lifting progress, cargo location, environmental parameters and other information in real time, and the operator can manually intervene at any time (such as suspending the lifting, emergency braking, etc.) to deal with emergencies; after the lifting is completed, the system records the data of this lifting operation for subsequent analysis and optimization.

Claims

1. A crane identification system based on artificial intelligence, characterized in that: include: Central control subsystem, as well as artificial intelligence cargo identification subsystem, intelligent lifting path planning subsystem, working environment perception and adaptation subsystem, human-computer interaction subsystem and displacement sensing subsystem connected to the central control subsystem; The artificial intelligence cargo identification subsystem is used to identify the appearance characteristics, size and weight of the cargo using a deep learning algorithm and a high-resolution image sensor to obtain complete cargo information; The intelligent lifting path planning subsystem is used to perform path simulation and planning based on cargo information using reinforcement learning to obtain the optimal lifting path; The working environment perception and adaptation subsystem is used to measure the ambient temperature, humidity and wind speed information using preset sensors; The human-machine interaction subsystem is used to utilize the human-machine interaction interface to exchange information between the operator and the crane identification system, and to manually intervene in the lifting operation to obtain human-machine interaction information; The displacement sensing subsystem is used to monitor the displacement information of the moving parts in real time by installing the displacement sensors on the moving parts of the crane; The central control subsystem is used to receive and process data from each subsystem, and issue control instructions to the crane's actuator according to the processed data.

2. The crane identification system based on artificial intelligence according to claim 1 is characterized in that: The central control subsystem receives and processes the data of each subsystem specifically as follows: Receive cargo information from the artificial intelligence cargo identification subsystem and coordinate the lifting process based on the cargo information; Receive the path instruction of the intelligent lifting path planning subsystem and convert the path instruction into a control signal for the actuator; Receive environmental data from the working environment perception and adaptation subsystem and adjust the crane operating parameters according to preset thresholds; Receive the optimal lifting path information planned by the intelligent lifting path planning subsystem and generate lifting instructions; Receive displacement information fed back by the displacement sensor system and control the lifting operation; The central control subsystem issues control instructions to the crane's actuator based on cargo information, actuator control signals, crane operating parameters, lifting instructions, and displacement information, driving the crane to move along the planned path.

3. The crane identification system based on artificial intelligence according to claim 2 is characterized in that: The artificial intelligence cargo identification subsystem includes: An image acquisition module is used to obtain image information of the cargo by placing a high-definition camera around the crane hook; An image recognition module is used to process the acquired cargo image information using a convolutional neural network to identify the type, shape, and size of the cargo; The weight detection module is used to measure the weight of the cargo by connecting a high-precision weighing sensor to the crane hook; The data integration module is used to integrate the type, shape, size information and weight of the object to generate complete cargo characteristic information.

4. The crane identification system based on artificial intelligence according to claim 3 is characterized in that: The intelligent lifting path planning subsystem includes: an environment detection module and a path planning and optimization module; The environment detection module is used to obtain the environment information around the crane using laser radar and ultrasonic sensors; The path planning and optimization module is based on a preset artificial intelligence algorithm and is used to combine three-dimensional environmental information, close-range obstacle information, complete cargo feature information and the performance parameters of the crane itself. During the lifting process, the lifting path is adjusted in real time according to environmental changes to generate the optimal lifting path.

5. The artificial intelligence-based crane identification system according to claim 4, characterized in that: The environment detection module specifically includes: The radar scanning submodule is used to scan the working space and obtain three-dimensional environmental information by installing the laser radar on the road working area of ​​the crane; The ultrasonic detection submodule uses ultrasonic waves to obtain obstacle information at close range by installing ultrasonic sensors on the hook, boom or body of the crane.

6. The artificial intelligence-based crane identification system according to claim 5, characterized in that: The execution steps of the preset artificial intelligence algorithm used in the path planning and optimization module are as follows: Based on 3D environmental information, close-range obstacle information, complete cargo feature information, and the crane’s own performance parameters; Initialize the reinforcement learning algorithm, set the state space of the reinforcement learning algorithm, and use the three-dimensional environment information, close-range obstacle information, complete cargo feature information, and the performance parameters of the crane itself as variables in the state space; Define the action space and speed adjustment range, and set the reward function based on the lifting path length, collision risk, and lifting efficiency; In the initial state, an action is selected from the action space according to the current crane state information, and the selected action is executed. The reward value is calculated according to the changed environment state and the set reward function; Adjust the strategy based on reward feedback, conduct experiments and learn in a simulated environment, and after multiple iterations, learn the optimal strategy in various situations; The lifting path is adjusted in real time based on the learned optimal strategy, the current crane parameters, and the dynamic changes in the environment to generate the optimal lifting path from the initial location of the cargo to the target location.

7. The crane identification system based on artificial intelligence according to claim 6 is characterized in that: The working environment perception and adaptation subsystem includes: an environment monitoring module and a parameter adjustment module; The environment monitoring module is used to detect environmental information of the working environment using an environmental sensor; The parameter adjustment module is used to adjust the operating parameters of the crane according to the ambient temperature, humidity and wind speed information and preset thresholds of the corresponding environmental information.

8. The artificial intelligence-based crane identification system according to claim 7, characterized in that: The environmental monitoring module specifically includes: The temperature monitoring submodule is used to monitor the temperature information of the working environment in real time by installing temperature sensors around the motor and electrical control cabinet; The humidity monitoring submodule is used to monitor the humidity information of the working environment in real time by installing humidity sensors in the concentrated area of ​​electrical equipment; The wind speed monitoring submodule is used to monitor the wind speed information of the working environment in real time by installing a wind speed sensor on the top of the crane.

9. The crane identification system based on artificial intelligence according to claim 1, characterized in that: The human-computer interaction subsystem includes: The information display module provides a display screen for displaying cargo information, lifting path, working environment parameters and system operation status information; The operation input module provides operation buttons and a touch screen to input manual lifting task instructions, manually intervene in lifting operations, and feed back human-computer interaction information to the central control subsystem.

Images