Tower crane control method and system based on auxiliary driving and terminal equipment

Abstract

The embodiment of the invention provides a tower crane control method and system based on auxiliary driving and terminal equipment, and belongs to the technical field of tower crane operation control. The method comprises the following steps: deploying a control system in tower crane equipment; according to the historical state information and historical mode selection information of the tower crane equipment and the historical execution condition of the tower crane operation task, an A3C algorithm is adopted to construct a self-adaptive decision model of the tower crane equipment, and the self-adaptive decision model is used for selecting an auxiliary driving mode most matched with the tower crane equipment; acquiring real-time state information of the tower crane equipment, and configuring a target mode matched with the real-time state information by adopting a self-adaptive decision model; receiving an operation instruction sent by a user to the tower crane equipment, and constructing a tower crane task model of the tower crane equipment in the target mode based on the operation instruction; displaying the tower crane task model to the user; and after the user confirms the tower crane task model, the tower crane task model is issued to the control system so as to control the tower crane equipment to complete the corresponding operation task according to the operation time sequence.

Description

Technical Field

[0001] The present application relates to the technical field of tower crane operation control, and in particular to a tower crane control method, system and terminal device based on assisted driving. Background Art

[0002] The traditional tower crane control method is mainly manual operation by the tower crane driver in the cab.

[0003] The tower crane driver operates the joysticks, buttons and other control devices in the cab to achieve actions such as amplitude change, lifting, and rotation. These devices are connected to actuators such as motors and hydraulic systems, and the driver manually adjusts the tower crane according to the lifting needs. During the operation, the driver needs to judge the lifting parameters and control the tower crane based on the lifting, hook dropping and other instructions conveyed by the ground commander through gestures, intercoms, etc. This control method requires high operating skills and experience of the driver. In complex working conditions such as the coordination of multiple tower cranes or in a small space, it is easy for the driver to make operating errors due to lack of experience and cause safety accidents. At the same time, due to the limited field of vision of the driver, it may be difficult to accurately judge the lifting position and posture for some complex lifting tasks, affecting work efficiency. Summary of the invention

[0004] The main purpose of the embodiments of the present application is to provide a tower crane control method, system and terminal device based on assisted driving, aiming to solve at least one technical problem existing in the traditional tower crane control method.

[0005] In a first aspect, an embodiment of the present application provides a tower crane control method based on assisted driving, comprising:

[0006] A control system is deployed in the tower crane equipment; the control system is connected to the linkage platform, and a variety of assisted driving modes are integrated in the control system; the various assisted driving modes are used to assist users to adopt corresponding operation modes to realize the control operation of the tower crane equipment; the various assisted driving modes at least include: linkage platform mode, voice driving mode, fixed distance driving mode, micro distance control mode, and somatosensory control mode; according to the historical status information of the tower crane equipment, the historical mode selection information, and the historical execution of the tower crane operation task, the A3C algorithm is used to build an adaptive decision model of the tower crane equipment to select the assisted driving mode that best matches the tower crane equipment; the real-time status information of the tower crane equipment is obtained, and the assisted driving mode is used The adaptive decision model configures a target mode that matches the real-time status information; receives operation instructions issued by the user to the tower crane equipment, and constructs a tower crane task model of the tower crane equipment in the target mode based on the operation instructions; a plurality of task nodes are arranged in the tower crane task model in order from first to last according to the operation sequence, and each task node corresponds to an operation task to be executed; the operation instruction is at least one of the linkage platform operation instruction, voice instruction, fixed-distance driving instruction, micro-distance control instruction, and somatosensory control instruction; the tower crane task model is displayed to the user; after the user confirms the tower crane task model, the tower crane task model is sent to the control system to control the tower crane equipment to complete the corresponding operation task according to the operation sequence.

[0007] In a second aspect, an embodiment of the present application provides a tower crane control system based on assisted driving, including:

[0008] A deployment module is used to deploy a control system in a tower crane. The control system is connected to a linkage platform, and a plurality of assisted driving modes are integrated in the control system. The plurality of assisted driving modes are used to assist the user to adopt corresponding operation modes to realize control operation of the tower crane. The plurality of assisted driving modes at least include linkage platform mode, voice driving mode, fixed-distance driving mode, micro-distance control mode, and somatosensory control mode. A decision module is used to construct an adaptive decision model of the tower crane using the A3C algorithm according to the historical status information of the tower crane, the historical mode selection information, and the historical execution status of the tower crane operation task, so as to select the assisted driving mode that best matches the tower crane. A configuration module is used to obtain the real-time status information of the tower crane and adopt an automatic A target mode that adapts to the configuration of the decision model and matches the real-time status information; a task module, which is used to receive the operation instructions issued by the user to the tower crane equipment, and build a tower crane task model of the tower crane equipment in the target mode based on the operation instructions; a plurality of task nodes are arranged in the tower crane task model in order from first to last according to the operation sequence, and each task node corresponds to an operation task to be executed; the operation instruction is at least one of the linkage table operation instruction, voice instruction, fixed-distance driving instruction, micro-distance control instruction, and somatosensory control instruction; a display module, which is used to display the tower crane task model to the user; a sending module, which is used to send the tower crane task model to the control system after the user confirms the tower crane task model, so as to control the tower crane equipment to complete the corresponding operation task according to the operation sequence.

[0009] In the third aspect, an embodiment of the present application also provides a terminal device, which includes a processor and a memory for storing a computer program; the processor is used to execute the computer program and implement the tower crane control method based on assisted driving as described in the first aspect or any embodiment of the present application when executing the computer program.

[0010] The embodiment of the present application provides a tower crane control method, system and terminal device based on assisted driving. The method includes: first, deploying a control system connected to a linkage platform in the tower crane equipment, integrating a linkage platform, voice, fixed distance, macro distance, somatosensory and other assisted driving modes to meet different working conditions and operator requirements. For example, those with insufficient experience can select the distance mode to reduce the difficulty of operation, which is convenient for operators to switch modes flexibly. The A3C algorithm is adopted to build an adaptive decision model based on the historical status, mode selection and operation task execution of the tower crane equipment, automatically recommend the most matching assisted driving mode, assist in reducing mode selection errors, and the model is optimized with new data. The real-time status information of the tower crane is obtained, and the adaptive decision model quickly configures the matching target mode. Under complex working conditions, the mode is adjusted according to the real-time distance, obstacle position, etc. to avoid safety accidents and ensure the stable operation of the tower crane. Finally, the user's operation instructions are received, the tower crane task model under the target mode is built and displayed, and sent to the control system after the user confirms. The task model sets multiple task nodes according to the operation sequence, provides clear operation guidance, and assists operators to understand the task process and requirements in advance, reduce operation delays, and improve efficiency. Issuing after confirmation can increase the field of vision and improve operational safety, accuracy and reliability. BRIEF DESCRIPTION OF THE DRAWINGS

[0011] Figure 1 A flowchart of a tower crane control method based on assisted driving provided in an embodiment of the present application;

[0012] Figure 2 A schematic diagram of the module structure of a tower crane control system based on assisted driving provided in an embodiment of the present application;

[0013] Figure 3 A schematic block diagram of the structure of a terminal device provided in an embodiment of the present application. DETAILED DESCRIPTION

[0014] In view of the technical problems existing in the traditional tower crane control method, the embodiment of the present application proposes a tower crane control method, system and terminal device based on assisted driving. Specifically, in view of the high requirements for driver operation skills, easy mistakes in complex working conditions, and limited vision affecting efficiency, the embodiment of the present application realizes efficient and safe operation through multi-mode integration and intelligent decision-making. The control system integrates multiple assisted driving modes such as linkage platform, voice, fixed distance, macro distance, and somatosensory. The driver can flexibly choose according to the task requirements, such as using the fixed distance mode for simple lifting to reduce the difficulty of operation. At the same time, the adaptive decision model based on the A3C algorithm can automatically match the optimal mode according to the historical information of the tower crane, and assist inexperienced drivers to reduce mistakes. Under complex working conditions, the system obtains equipment status information in real time and dynamically selects the adaptation mode, such as switching to the macro distance control mode to avoid collision risks when obstacles are detected, and plans the operation sequence by building a tower crane task model to ensure orderly operation. In addition, it supports multiple operation instructions, combines the somatosensory control mode of high-definition cameras and AR devices to broaden the driver's field of vision, and cooperates with the visual display and confirmation mechanism of the tower crane task model to help the driver master the task process in advance, which not only improves the work efficiency, but also ensures the safety of operation.

[0015] The embodiment of the present application provides a tower crane control method, system and terminal device based on assisted driving. Among them, the tower crane control method based on assisted driving can be applied to the terminal device, and the terminal device can be a linkage station in the tower crane device, or a linkage station in the tower crane device, and a mobile terminal communicating with the linkage station, such as a mobile phone, a virtual reality device, a tablet computer, a laptop computer, a desktop computer, a wearable device and other electronic devices. The terminal device can be a server connected to the tower crane device, or it can be a server cluster. The above connection method can be implemented through a hardware circuit or a communication module.

[0016] Some embodiments of the present application are described in detail below in conjunction with the accompanying drawings. In the absence of conflict, the following embodiments and features in the embodiments can be combined with each other. Figure 1 , Figure 1 A flowchart of a tower crane control method based on assisted driving is provided in an embodiment of the present application.

[0017] like Figure 1 As shown, the tower crane control method based on assisted driving includes steps S101 to S106.

[0018] Step S101: deploy a control system in a tower crane.

[0019] In the embodiment of the present application, the control system is the core part of the intelligent operation of the tower crane. The control system realizes the comprehensive management and control of the tower crane through hardware devices (such as sensors, controllers, communication modules, etc.) and software programs. For example, in terms of hardware, the sensor collects the operation data of the tower crane in real time, such as the hook position, load weight, boom angle, etc. The controller analyzes and processes these data. The communication module is responsible for data transmission with other devices (such as linkage stations, remote monitoring centers, etc.). The software program includes various control algorithms, decision logics and operation interfaces, etc., which can realize real-time monitoring, fault diagnosis, safety protection and intelligent control of the tower crane's operating status.

[0020] For example, the control system is connected to the linkage platform, and a variety of assisted driving modes are integrated in the control system. The control system is connected to the linkage platform, which not only supports traditional manual operation, but also integrates a variety of advanced assisted driving modes, so that the tower crane has more flexible and intelligent operation capabilities.

[0021] The linkage platform is the core hardware equipment for tower crane operation. It can be understood as the interactive hub between the tower crane driver and the mechanical action of the tower crane, and is a core component that can convert the driver's operating intention into the control instructions for the actual operation of the tower crane. For example, the linkage platform can be installed in the tower crane cab and consists of multiple joysticks, buttons, switches, indicator lights, somatosensory control components, and multiple sensors or data acquisition modules that match the assisted driving mode.

[0022] In the embodiment of the present application, a plurality of auxiliary driving modes are used to assist the user to adopt the corresponding operation mode to realize the control operation of the tower crane equipment, aiming to assist the user to operate the tower crane more efficiently and safely from different dimensions, so as to comprehensively improve the intelligence and safety of the tower crane operation.

[0023] Exemplarily, the multiple assisted driving modes include at least: linkage table mode, voice driving mode, fixed-distance driving mode, macro control mode, and somatosensory control mode. Each mode is designed for different operating scenarios and requirements. In linkage table mode, the linkage table handle and button are used as the core of operation to accurately transmit the commands such as lifting the boom and moving the hook to the control system, retain the traditional operation characteristics, easy to use, and standardized process, which is suitable for common operating scenarios. In voice driving mode, the voice recognition module converts the command such as "lift the hook 5 meters" into text, executes it after parsing, and verifies the identity in combination with voiceprint recognition to ensure safe operation. This mode frees your hands and is especially suitable for scenes where your hands are busy or space is limited. In fixed-distance driving mode, after entering the target distance, the system automatically calculates the hook movement parameters, calibrates in real time with the help of laser radar or visual sensors, and accurately controls the hook position. It is suitable for operations with high precision requirements such as precise placement of building components. In macro control mode, the command is input through the knob or touch screen to achieve centimeter-level fine movement of the hook, and feedback the position information in real time, meeting the needs of high-precision installation operations at close range and reducing operational risks. In the somatosensory control mode, the high-definition camera or AR device under the tower is used to collect images. The operator directly operates the tower crane through a touch screen or AR glasses, breaking the communication barriers between the tower and the tower, providing a wide field of view, and significantly improving the operation efficiency and safety in a complex environment. It is understandable that in the related technology, the traditional linkage platform is mainly used to control the basic actions of the tower crane, such as lifting and lowering the boom, moving the hook, rotating, etc. In the embodiment of the present application, in order to cooperate with a variety of auxiliary driving modes, the linkage platform integrates more functions. For example, in terms of coordination with the voice driving mode, the linkage platform has added a voice command receiving and parsing module, which can convert the instructions parsed by the voice recognition module into corresponding control signals to achieve precise control of the tower crane. After receiving and recognizing the voice command of "lifting the hook 5 meters", the linkage platform quickly converts this command into a signal to control the operation of the hook lifting motor, so that the hook rises 5 meters accurately. For the fixed-distance driving mode, the linkage platform integrates the target distance input interface and the functional modules related to the automatic control algorithm. After the operator enters the target distance, the linkage platform can automatically adjust the hook movement speed and distance according to the built-in algorithm to ensure that the tower crane reaches the specified position accurately.

[0024] In terms of optimizing the convenience of operation, the linkage console re-arranges key components such as handles and buttons, and arranges them according to the frequency and relevance of operation to reduce the operation range and switching time. At the same time, it is centrally arranged according to the operation logic, combined with a variety of assisted driving modes to achieve linkage control, and simplify complex processes with one-touch buttons. The operating components are clearly marked and distinguished by touch to improve the accuracy and efficiency of operation in poor light or emergency situations.

[0025] In terms of security protection, the operation permission management function has been added. By integrating with biometric technologies such as voiceprint recognition and fingerprint recognition, only authorized operators can operate the linkage table to prevent unauthorized personnel from misoperating and causing safety accidents. Redundant design and fault isolation technology are used inside the linkage table. When a circuit module fails, it will not affect the normal operation of other modules, ensuring the continuity and stability of tower crane operation.

[0026] In order to better integrate with other assisted driving modes and realize intelligent interaction, the linkage platform is equipped with an intelligent display screen. The display screen can display the operating status of the tower crane, the execution of operating instructions, the working parameters of various assisted driving modes and other information in real time. In the macro control mode, the display screen can accurately display the centimeter-level moving position of the hook, providing accurate feedback to the operator. The linkage platform has a communication interface with other intelligent devices or systems, which can realize remote monitoring and operation. Managers can connect to the linkage platform through terminal devices such as mobile phones and computers to understand the operation status of the tower crane in real time. If necessary, they can also remotely operate the tower crane to improve management efficiency and emergency response capabilities. In the group tower operation scene, the linkage platform can exchange information with the linkage platforms of other tower cranes, plan the operation path of the tower crane through intelligent algorithms, and avoid collision accidents.

[0027] Step S102: Based on the historical status information, historical mode selection information, and historical execution status of the tower crane operation tasks, an A3C algorithm is used to construct an adaptive decision model for the tower crane to select the assisted driving mode that best matches the tower crane.

[0028] Step S103: acquiring real-time status information of the tower crane equipment, and using an adaptive decision model to configure a target mode that matches the real-time status information.

[0029] In the embodiment of the present application, the adaptive decision model is a model constructed based on the A3C (Asynchronous Advantage Actor-Critic) algorithm in deep reinforcement learning (DRL). The model can adaptively select the most suitable assisted driving mode for the current working conditions based on the historical data and real-time status information of the tower crane. The core of this model is to automatically adjust the strategy by continuously learning the relationship between mode selection and task execution results in historical data to achieve dynamic optimization of the operation mode.

[0030] In step S102, the A3C algorithm is a model-free deep reinforcement learning algorithm that combines policy gradient (Actor) and value function estimation (Critic). When building an adaptive decision model, the historical state information of the tower crane (such as hook position, boom angle, load weight, etc.), historical mode selection information (assisted driving modes used in the past), and historical execution of tower crane tasks (whether the task was successfully completed, completion time, etc.) are used as training data. For example, through multi-threaded asynchronous training, multiple agents simultaneously learn and explore in different copies of the environment, and continuously update the parameters of the Actor network (used to generate action probability distribution) and the Critic network (used to evaluate state value), so that the model can learn the strategy of selecting the optimal operation mode under different states.

[0031] For example, assume that there is historical data of the past 1,000 tower crane operations, which contains the state information of the tower crane during each operation (such as hook height, load weight, etc.), the selected assisted driving mode (linkage mode, voice driving mode, etc.), and the execution results of the operation (successful completion of the task, task timeout, etc.). Divide this data into training sets and validation sets, and use the A3C algorithm for training. During the training process, the Actor network will output the probability of selecting different assisted driving modes based on the current state, and the Critic network will evaluate the value of the current state. By continuously adjusting the network parameters, the model can select the assisted driving mode that is most likely to lead to successful completion of the task under different states.

[0032] Therefore, by learning from historical data, the model can better understand the applicability of various assisted driving modes under different conditions, and assist in the automation and intelligence of operating mode selection. As historical data continues to accumulate, the model can be continuously updated and optimized to adapt to different working conditions and task requirements.

[0033] Further optionally, in step S102, multiple assisted driving modes are defined as action spaces, and an assisted driving mode is selected from the action space in each time step; a reward function of an adaptive decision model is set based on historical execution conditions; historical state information, historical mode selection information, and historical execution conditions of tower crane operation tasks are obtained; multiple operation task environments are constructed based on historical state information; historical state information and historical mode selection information are input into multiple operation task environments, and the Actor network used to select the assisted driving mode is trained in parallel; historical state information, historical mode selection information, and historical execution conditions are input into multiple operation task environments, and the Critic network used to evaluate the value of the Actor network is used in parallel; the reward function is used to update the network parameters of the Actor network and the Critic network to obtain an adaptive decision model for the tower crane. In the embodiment of the present application, the reward function includes at least: task completion reward, completion time reward, completion quality reward, and abnormal situation penalty for the tower crane.

[0034] Specifically, in the tower crane assisted driving scenario, multiple assisted driving modes such as linkage mode, voice driving mode, fixed-distance driving mode, micro-distance control mode, and somatosensory control mode are defined as action space. At each time step, the model needs to select an assisted driving mode from this action space to execute. For example, at a certain moment, the model must decide whether to use voice driving mode or fixed-distance driving mode to control the tower crane. Furthermore, the reward function is set based on the historical execution of the tower crane operation task. If the tower crane equipment successfully completes the lifting task, a positive reward is given as a task completion reward. For example, if the material is accurately lifted to the specified location, a higher reward value can be obtained; conversely, if the task fails, a penalty is given. Quick completion of the task is encouraged. If the actual completion time of the task is shorter than the expected time, an additional reward is given; if it exceeds the expected time, a penalty is given. When the tower crane encounters abnormal conditions, such as collision with obstacles, overload operation, etc., corresponding penalties are given to avoid such situations from happening again. Next, collect the historical status information of the tower crane equipment, such as hook position, boom angle, load weight, etc.; historical mode selection information, that is, the assisted driving mode used in the past; and the historical execution of the tower crane operation task, such as whether the task was successful, completion time, completion quality, etc. These historical data are the basis for building and training the model. Then, build multiple operation task environments based on the historical status information. Each environment simulates different tower crane operation scenarios, such as different load weights, different operation spaces, different weather conditions, etc. This allows model training in multiple scenarios to improve the generalization ability of the model. On the one hand, the historical status information and historical mode selection information are input into multiple operation task environments, and the Actor network used to select the assisted driving mode is trained in parallel. The role of the Actor network is to output the probability distribution of selecting each assisted driving mode based on the current state information. By continuously training in different operation task environments, the Actor network can learn which assisted driving mode should be selected in different states to obtain the maximum reward. On the other hand, the historical status information, historical mode selection information, and historical execution status can be input into multiple operation task environments, and the Critic network used to evaluate the value of the Actor network can be trained in parallel. The main function of the Critic network is to evaluate the value of the action (assisted driving mode) selected by the Actor network in the current state. It will judge whether the current action is reasonable based on the reward value given by the reward function, and provide feedback for the training of the Actor network. Finally, the reward function can be used to update the network parameters of the Actor network and the Critic network. During the training process, the model will continuously adjust the network parameters based on the feedback of the reward function, so that the Actor network can more accurately select the assisted driving mode and the Critic network can more accurately evaluate the value of the action. After multiple iterations of training, the adaptive decision model of the tower crane equipment is finally obtained.

[0035] It is understandable that, compared with the single-target training in the traditional A3C algorithm, in the embodiment of the present application, in the tower crane assisted driving scenario, the reward function comprehensively considers multiple factors such as task completion, completion time, completion quality and abnormal situations, which is more in line with the actual needs of tower crane operations. For example, a specific design is performed based on the tower crane scenario, and hard safety constraints (abnormal situation penalties) are added to give priority to avoiding equipment damage or accidents, which meets the core needs of industrial control. Alternatively, the task quality rewards (accuracy, time) are subdivided to match the actual needs of "precise lifting" and "efficient construction" in tower crane operations. At the same time, mode switching penalties are introduced to suppress strategy shocks and improve operational stability (traditional algorithms may cause frequent mode switching due to excessive exploration). This can guide the model to focus on efficiency and quality while ensuring task completion, avoid abnormal situations, and improve the practicality and safety of the model.

[0036] The traditional A3C algorithm has limitations in the processing of state space and action space. It usually only targets continuous actions or simple discrete actions (such as button operations in games), and the state space is mostly visual images or low-dimensional vectors, which makes it difficult to adapt to the complex needs of tower crane assisted driving scenarios.

[0037] In the embodiment of the present application, the adaptive decision model in the tower crane assisted driving scenario further improves its adaptability to working conditions through engineering adaptation. Specifically, in the construction of the state space, multimodal industrial data is integrated, including equipment status data such as hook position, load weight, crane arm angle, etc. collected by sensors in real time, task characteristics such as target position, accuracy requirements, time limit in task attributes, and environmental parameters such as wind speed, obstacle distribution, and light intensity. Through feature engineering, key dimensions such as load weight and wind speed are normalized to form a high-dimensional state vector with both physical meaning, so that the model can accurately capture the complex working conditions of tower crane operation.

[0038] Optionally, a self-supervised learning algorithm (such as SimCLRv2) can be introduced to enhance the sensor time series data, automatically mine the mechanical fatigue characteristics in the load fluctuation signal, the risk association after continuous emergency stop and other implicit information, and combine the Gaussian process regression model to predict the trend of environmental changes, so that the state vector is upgraded from a simple physical parameter splicing to an intelligent representation that includes time series dependency and risk prediction. Through feature engineering processing such as normalization of load weight percentage and wind speed Z-score standardization, the physical meaning of the data is further enhanced. At the same time, digital twin technology is used to build a high-precision dynamic simulation environment to generate samples of extreme working conditions (such as ultra-limit wind speed, load surge) missing in historical data, and supplement the dangerous scene coverage of the training data, so that the model can accurately capture the coupling effect of the boom angle and wind speed, the dynamic effect of load inertia on the swing of the hook and other complex working conditions.

[0039] In terms of action space design, we break through the direct processing of low-level control actions such as motor speed and joint angle by traditional algorithms, and abstract them into discrete assisted driving mode selection. Each mode encapsulates mature upper-level control strategies. For example, the voice mode integrates the semantic parsing model based on Transformer to realize multi-command context understanding, and the fixed-distance mode embeds the PID parameter self-tuning algorithm combined with visual feedback to improve positioning accuracy. On this basis, the meta-learning (MAML) technology is introduced to initialize the strategy parameters of different task types (heavy-load lifting, precision positioning), so that the model can quickly adapt to the dynamic parameter changes of the new model tower crane through a small number of samples. At the same time, a three-level redundant decision-making mechanism of the main strategy (A3C generation mode), backup strategy (heuristic rules), and expert rules (safety bottom line) is constructed, and the decision results are integrated using the Dempster-Shafer evidence theory. When critical working conditions such as wind speed>20m / s are detected, the coordinated control of the wind resistance mode and the safety fallback strategy is automatically triggered. In this way, the adaptive ability of reinforcement learning is retained, and the model is prevented from outputting dangerous decisions through hard-coded safety rules. This abstract processing and policy-level encapsulation reduces the action space dimension from dozens of underlying control parameters of traditional algorithms to five mode choices. While reducing the number of model parameters, it also improves decision reliability, significantly reduces the difficulty of engineering deployment and enhances adaptability to actual scenarios.

[0040] For example, in a tower crane operation at a bridge construction site, the improved state space construction technology significantly improved the model's ability to understand complex working conditions through multi-dimensional data fusion and intelligent processing. When the tower crane continuously lifted a 50-ton box girder, the load fluctuation signal collected by the sensor was processed by SimCLRv2 self-supervised learning, and the periodic load abnormal fluctuation characteristics caused by slight wear of the wire rope were automatically identified. This feature is easily overlooked in traditional artificial feature engineering, and the adaptive decision model found through comparative learning that it has an 89% correlation with the subsequent wire rope breakage accident within 2 hours, thereby triggering an equipment maintenance warning in advance.

[0041] At the same time, the adaptive decision model can also be processed through Gaussian process regression. Based on historical wind speed data and real-time weather radar signals, it is predicted that a strong wind of 18m / s will be encountered in the next 30 minutes. The system automatically integrates parameters such as "wind speed change trend", "current angle of crane arm" and "load inertia moment" into the state vector dynamically, so that the model can be pre-adjusted to wind resistance mode 10 minutes before the strong wind arrives. Further, optionally, the composite working conditions of super-limit wind speed and full load that are missing from historical data can be generated through digital twins to simulate the following scenarios: simulating the critical instability state of the crane arm at a wind speed of 25m / s, the adaptive decision model learns from the virtual data that "when the product of wind speed and crane arm angle exceeds the safety threshold, unloading in stages is required even if the load limit is not reached", filling the gap in decision-making for dangerous working conditions that is difficult to verify in real scenarios.

[0042] When the operator issues a voice command "Lift the prefabricated panel to a height of 20 meters, move it to the coordinates (15, 8, 20) and then fine-tune the placement", the voice pattern analysis module integrated with Transformer not only recognizes the single command, but also captures the precise positioning intention behind this "fine-tuning" command, and automatically combines the fixed-distance mode and the macro-distance control mode. In this way, the moving path can be controlled with an accuracy of 1cm in the fixed-distance mode through the PID algorithm embedded with visual feedback, and the vibration compensation algorithm is enabled in the last 50cm of the macro mode, reducing the component placement error from ±5cm of the traditional solution to ±1.2cm.

[0043] When faced with a new tower crane, assuming that the boom length is 15% longer than the old model, the adaptive decision model can also use the MAML algorithm to quickly adjust the path planning parameters of the fixed-distance mode based on the boom angle and speed data collected by the new equipment during three trial lifts. This reduces the first task completion time of the new equipment from 40 minutes using traditional transfer learning to 12 minutes, while maintaining the same positioning accuracy.

[0044] In the actual application of the three-level redundant decision-making mechanism, when the tower crane performs heavy-load lifting at a wind speed of 16m / s, the main strategy tends to choose the linkage mode because there are fewer samples in this wind speed range in the historical data. However, its risk assessment score (0.65) is lower than the heuristic rule score of the backup strategy (0.82, rule: automatic mode is preferred when wind speed>15m / s). Combined with the safety bottom line of the expert rule (manual high-speed rotation is prohibited when wind speed>12m / s), the adaptive decision model finally triggers the coordinated control of the wind-resistant mode and the fixed-distance mode, limiting the crane arm rotation speed to 0.3° / s.

[0045] Taking the tower crane hoisting prefabricated beams of different weights (such as 70-ton T-beams and 30-ton hollow slabs) as an example, each assisted driving mode encapsulates an independent control strategy. The fixed-distance mode is equipped with a load-based PID parameter library. For example, the PID parameters corresponding to a 70-ton load are: P=0.8, I=0.3, and D=0.5. The wind-resistant mode dynamically adjusts the boom elevation angle according to the wind speed, gravity acceleration, and boom length through a formula. The wind speed is collected in real time by the sensor, the gravity acceleration is about 9.8m / s² as a constant, and the boom length is determined according to the tower crane model and working conditions. By adjusting the elevation angle to change the force distribution of the tower crane, increasing the elevation angle in strong winds can reduce lateral wind pressure, avoid overturning, and ensure the stability of the tower crane.

[0046] Meta-learning can play a role in strategy acceleration in this architecture. First, the meta-learner extracts the meta-parameters of each mode from historical lifting tasks. Take the task of "for every 10 tons increase in load, the PID-P parameter of the fixed distance mode increases by an average of 0.15" as an example, and encode it into meta-knowledge. When the new model tower crane (the boom length L increased from 45 meters to 55 meters) performed the 80-ton box girder lifting for the first time, the meta-learner, based on the encapsulated wind-resistant mode strategy framework, used the "mapping relationship between L and θ" meta-knowledge in historical tasks to directly generate the initial control parameters (referring to the above formula, it can be obtained that θ is approximately equal to 19.7°), which reduces the number of trial and error by 70% compared with traditional random initialization.

[0047] In more complex composite working conditions (such as the "strong wind + variable load" scenario), the mode-encapsulated strategy unit becomes the basic operation unit of meta-learning. When the wind speed suddenly increases from 12m / s to 18m / s and the load is dynamically adjusted from 60 tons to 40 tons, the meta-learning algorithm (such as MAML) no longer optimizes the underlying motor control parameters, but instead fine-tunes the joint parameters of the encapsulated wind-resistant mode and fixed-distance mode, that is, through 10 virtual environment trainings (taking less than 2 seconds), it is quickly determined that the boom angle compensation coefficient of the wind-resistant mode is adjusted from 0.9 to 1.2, and the path planning speed threshold of the fixed-distance mode is increased from 0.5m / s to 0.7m / s, which increases the model's first decision accuracy under new working conditions from 30% to 85%. This combination essentially transforms the domain knowledge encapsulated in the mode into a transferable strategy prior for meta-learning. It not only retains the professionalism of each mode (such as the physical model constraints of the wind-resistant mode), but also realizes the knowledge transfer between modes through meta-learning (such as the correlation rules of mode parameters under different loads). Ultimately, in a certain construction project, the strategy adaptation time of new equipment was shortened from 48 hours to 1.5 hours, and the mode switching failure rate under complex working conditions was reduced by 60%.

[0048] In this way, the policy parameters (such as PID coefficients and semantic parsing model weights) encapsulated in each mode constitute the subtasks of meta-learning. The meta-learner optimizes the initialization parameters of these subtasks through gradient descent, so that they can converge quickly in new tasks with only a small number of samples. For example, in the semantic parsing model of speech mode, meta-learning found that the command features related to the lifting accuracy requirements (such as keywords such as precision and fine-tuning) have commonality in the word vector mapping in different projects. Therefore, when deployed at a new construction site, fewer new command samples can be used to complete the adaptation of the parsing model, reducing the need for labeled data by 80% compared to traditional fine-tuning methods.

[0049] The above example avoids the complexity of directly operating the underlying control actions, and endows the model with policy-level migration capabilities through meta-learning, enabling the tower crane assisted driving system to demonstrate industrial-level rapid adaptability when facing equipment iterations and changes in working conditions.

[0050] In an embodiment of the present application, the adaptive decision model supports an online fine-tuning mechanism. When new operating condition data such as new model tower crane data and extreme weather operation data accumulate to a certain scale, the model parameters are updated through incremental training instead of retraining the entire network, which significantly improves the model's adaptability to new scenarios and training efficiency.

[0051] In response to the gradient oscillation problem caused by environmental differences in traditional A3C multi-threaded asynchronous training, the adaptive decision model achieves a dual improvement in training stability and efficiency through an engineered acceleration strategy. For example, in the embodiment of the present application, the adaptive decision model clusters environmental copies by task type, and divides complex tower crane operations into typical scenarios such as "heavy load lifting", "high-precision positioning", and "strong wind environment operations". Each worker thread focuses on strategy training under specific working conditions, reducing the interference caused by the mixing of data from different scenarios and improving sample utilization efficiency.

[0052] Furthermore, in the embodiments of the present application, a priority experience replay mechanism is introduced to assign higher weights to the trajectories of high-reward events (such as precise operations to successfully avoid obstacles) and high-risk events (such as safety control under critical overload conditions). By prioritizing the learning of these key experiences, the model's convergence to safe and efficient strategies is accelerated, avoiding the problem of dilution of important experiences due to random sampling in traditional algorithms, and enabling the training process to optimize key decision-making logic in a more targeted manner.

[0053] After the adaptive decision model is built, in step S103, the real-time status information of the tower crane is obtained, including the current hook position, boom angle, load weight, and environmental parameters (such as wind speed, light, etc.). These real-time status information are input into the trained adaptive decision model, and the model outputs the most suitable assisted driving mode for the current state, i.e., the target mode, according to its learned strategy.

[0054] For example, at a certain moment, the real-time status information of the tower crane is obtained: the hook height is 10 meters, the load weight is 5 tons, and the wind speed is 3m / s. This information is input into the adaptive decision model, and the model judges based on its internal strategy that the most suitable auxiliary driving mode is the fixed-distance driving mode, because in this state, the fixed-distance driving mode can more accurately control the movement of the hook to complete the current task. Thus, the target mode is dynamically selected according to the real-time status information to ensure that the tower crane can adopt the most suitable operation mode at any time, quickly adapt to different working conditions and adjust the operation mode in time, improve operation efficiency and safety, and ensure the stable operation of the tower crane.

[0055] As an optional embodiment, in step S103, an adaptive decision model is used to configure a target mode that matches the real-time status information, including: acquiring image data around the tower crane equipment, and locating obstacles based on the image data to obtain obstacle positioning data of the tower crane equipment; parsing the real-time status information to obtain the real-time equipment component status, real-time weather conditions, and real-time wind speed conditions of the tower crane equipment; inputting the real-time equipment component status, real-time weather conditions, real-time wind speed conditions, and obstacle positioning data into the adaptive decision model, using the Actor network to predict the driving mode, and obtaining the predicted probability of the tower crane equipment in each assisted driving mode; randomly selecting a candidate mode through roulette based on the predicted probability; monitoring the safety risk of the tower crane equipment in the candidate mode; if the safety risk is greater than the set threshold, setting the target mode to an assisted driving mode that matches the safety risk item, or switching the target mode to an assisted driving mode with a safety level higher than the current target mode.

[0056] In principle, first, the camera is used to collect image data around the device, and the image is processed through computer vision algorithms (such as YOLO, Mask R-CNN) to realize the identification and positioning of obstacles and obtain obstacle positioning data. At the same time, the real-time status information of the tower crane is analyzed, including the status of equipment components (such as hook position, boom angle), weather conditions, wind speed conditions, etc., to provide comprehensive data support for subsequent decision-making. Then the above data is input into the Actor network of the adaptive decision model, which predicts each assisted driving mode and obtains the probability of selecting each assisted driving mode. Based on the predicted probability output by the Actor network, the candidate mode is determined by roulette random selection. After the candidate mode is determined, the safety risk of the tower crane operation is monitored in real time. If the risk exceeds the threshold, the target mode is set to a mode that matches the safety risk according to the risk type, or switched to a higher safety level mode to ensure the safe operation of the tower crane.

[0057] For example, suppose that at a certain construction site, a tower crane is carrying out a lifting operation. In this case, the camera can capture some building materials and construction equipment around the tower crane. After locating the obstacles, it is determined that there is a pile of steel bars 3 meters away from the tower crane boom, which may hinder the rotation of the tower crane. Assume that the real-time equipment component status is that a 5-ton weight is hung on the hook, the boom angle is 45 degrees, and it is in the process of rotation. Assume that the real-time weather conditions are that it is about to rain. Assume that the real-time wind speed conditions are that the current wind speed is 12m / s and has a tendency to gradually increase. Based on these assumptions, after inputting this information into the Actor network of the adaptive decision model, the predicted probabilities of each assisted driving mode are obtained: the probability of the linkage mode is 0.3, the probability of the voice driving mode is 0.2, the probability of the fixed-distance driving mode is 0.25, the probability of the micro-distance control mode is 0.15, and the probability of the somatosensory control mode is 0.1. Through random selection by roulette, the candidate mode is determined to be the fixed-distance driving mode. When the tower crane was operating in fixed-distance driving mode, it was detected that as the wind speed continued to increase to 15m / s and the boom was close to obstacles, there was a high risk of collision and overturning, and the safety risk exceeded the set threshold. In this case, according to the safety risk situation, the target mode was switched to a combination of wind-resistant mode and micro-control mode: the wind-resistant mode is used to cope with strong winds and enhance the stability of the tower crane by adjusting the boom angle and operating parameters; the micro-control mode is used to perform precise micro-movement operations when approaching obstacles to avoid collisions and ensure the safe completion of the lifting operation.

[0058] In this way, the tower crane auxiliary driving system monitors safety risks in real time, dynamically adjusts the target mode, reduces the probability of accidents such as collisions and overturning, and ensures the safety of personnel and equipment. It integrates multi-source data to predict driving modes, flexibly selects modes based on the environment, equipment status and weather, and improves adaptability to different scenarios. The use of roulette random selection combined with risk monitoring ensures reasonable decision-making, avoids conservative and single mode selection, and improves decision-making and operation efficiency.

[0059] Step S104: receiving an operation instruction issued by a user to the tower crane equipment, and constructing a tower crane task model of the tower crane equipment in a target mode based on the operation instruction.

[0060] The tower crane task model is a model built for tower crane equipment to perform tasks based on user operation instructions and target patterns. The model can decompose and plan the entire operation process of the tower crane to ensure that the tower crane can complete the task accurately and safely according to the predetermined steps and requirements. According to the user's operation instructions and the current target pattern, the operation of the tower crane is planned as a whole, and the tasks to be completed at each stage are clarified, so that the operation of the tower crane has a clear process and sequence to avoid confusion and errors. The model coordinates the work between the various components and systems of the tower crane to ensure that at different task stages, the various components can work together as required to achieve precise position control, speed control and force control, etc., to complete specific tasks such as lifting cargo and adjusting the position of the boom. Through reasonable planning of tasks and strict control of each task node, the model can take into account various safety factors in advance, set up corresponding safety inspections and protection mechanisms, reduce safety risks during the operation, and ensure the safety of tower crane equipment and personnel. The model can be used to optimize the operation process, reduce unnecessary actions and time waste, improve the work efficiency of the tower crane, enable the tower crane to complete the task in the shortest time, and improve the construction progress.

[0061] Among them, in the tower crane task model, multiple task nodes are set in order from the first to the last according to the operation sequence, and each task node corresponds to a task to be executed. These task nodes are interrelated and executed in sequence, and together constitute the complete process of the tower crane completing an operation task. Each task node has clear task objectives and operation requirements. Through the precise control and coordination of these task nodes, it can be ensured that the tower crane completes various operation tasks safely and efficiently.

[0062] Exemplarily, in the tower crane task model, the data information that each task node can contain is as follows: In task node 1, in the preparation stage, check the status of the tower crane equipment components (hook, boom, wire rope, etc.), confirm whether the load weight exceeds the rated load, and check whether there are obstacles in the surrounding environment; at the same time, according to the target mode and operation instructions, initialize the tower crane control system, set the initial angle of the boom, the initial position of the hook and other parameters. In task node 2, in the lifting stage, according to the load weight and the target position, calculate and control the appropriate lifting speed and force, slowly lift the hook, and make the load leave the ground. During the lifting process, monitor the state of the load and the stability of the tower crane in real time to ensure a smooth lifting process and avoid load shaking or tilting of the tower crane. In task node 3, in the horizontal movement stage, after the load leaves the ground and reaches a certain height, control the boom to rotate and change the amplitude so that the load moves horizontally above the target position. During this process, it is necessary to accurately control the movement speed and angle of the boom according to factors such as wind speed, boom length and load weight, while maintaining the balance and stability of the load to prevent the load from swinging too much due to inertia or wind force. In task node 4, during the descent phase, when the load reaches above the target position, control the tower crane to slowly lower the hook height and accurately place the load at the specified position. During the descent process, closely monitor the distance and alignment between the load and the target position, and adjust the descent speed and hook position in time to ensure that the load can be placed smoothly and accurately. In task node 5, during the closing phase, after the load is placed, raise the hook and return it to its position, shut down the power system of the tower crane, inspect and maintain the tower crane equipment, and clean up the site. At the same time, record the relevant data of this operation, such as operation time, load weight, boom movement parameters, etc., for subsequent data analysis and equipment management.

[0063] Exemplarily, the operation instruction is at least one of the linkage platform operation instruction, voice instruction, fixed-distance driving instruction, micro-distance control instruction, and somatosensory control instruction. Specifically, the linkage platform operation instruction contains the action information of each mechanism of the tower crane (such as lifting, amplitude change, rotation, etc.), such as the speed and direction of the motor, and the action amplitude and position of each mechanism. It may also include the timestamp of the operation, which is used to record the time when the operation occurs, so as to perform sequential control and time-related analysis. The voice instruction contains the audio data of the voice, which is encoded and stored and transmitted in the form of digital signals. In addition, it also includes the parsed semantic information, such as specific operation commands (such as rising, falling, turning left, turning right, etc.), the target object of the operation (such as the lifted goods, the moving position, etc.), and possible modifiers (such as slow, fast, etc.), and may also include auxiliary information such as the source and time of the voice instruction. The fixed-distance driving instruction mainly contains distance-related data, such as the horizontal distance and vertical distance that the tower crane needs to move, and the coordinate information of the target position. It may also include speed information, that is, the speed at which the crane is expected to move to the specified distance, as well as acceleration limits and other data to ensure the stability and safety of the crane during movement, and information such as the execution time of the instruction. The macro control instruction contains relevant data for fine operations, such as precise control parameters of the motor, such as the adjustment values ​​of current and voltage, to achieve small movements. It also includes position feedback information, which is used to accurately monitor the current position and state of the crane for real-time adjustment. In addition, it may include data such as the accuracy requirements of the operation, the small offset of the target position, and time parameters related to macro operations, such as the duration and interval of the operation. The somatosensory control instruction contains posture data of human body movements, such as the shape, angle, position change and other information of the gesture, which are collected and converted into digital signals by sensors (such as cameras, somatosensory bracelets, etc.). It also includes time series information of the action, which is used to record the sequence and duration of gesture actions in order to learn the dynamic changes of gestures. In addition, it may include identification information related to the somatosensory device, as well as the start and end time of the operation.

[0064] As an optional embodiment, in step S104, a target parsing model matching the operation instruction is selected from the candidate algorithm model library; the operation instruction is processed using the target parsing model to obtain an operation instruction feature vector; the operation instruction feature vector contains the instruction sequence between each work task, the task type to which the work task belongs, and the logical relationship; based on the operation instruction feature vector and the correlation between the preset task type and task priority, the work tasks to be executed and the arrangement of the work tasks are determined; according to the work tasks to be executed and the arrangement of the work tasks, the tower crane task model is obtained through a graph neural network combination.

[0065] Specifically, the candidate algorithm model library stores a variety of parsing models for different operation instructions. According to the input operation instruction type, a matching target parsing model is selected from the library. For example, if it is a voice instruction, a model combining BERT, Transformer and self-attention mechanism is selected; if it is a somatosensory control instruction, a model combining DCNN and LSTM is selected. Then, the target parsing model is used to process the operation instruction and convert the operation instruction into an operation instruction feature vector. The feature vector contains important information such as the instruction sequence between each operation task, the task type to which the operation task belongs, and the logical relationship. For example, for a voice instruction containing multiple steps, "first lift the hook, then move it to the specified position, and finally lower the hook", the feature vector will record the order of these steps and the task type to which each step belongs (such as lifting, moving, lowering, etc.). Then, based on the operation instruction feature vector and the relationship between the pre-set task type and task priority, the operation tasks to be executed and the arrangement of the operation tasks are determined. Different task types may have different priorities. For example, safety-related tasks usually have higher priorities. According to these priorities and instruction sequence, the execution order and method of each operation task are determined. Finally, according to the tasks to be performed and the arrangement of the tasks, the tower crane task model is obtained through the combination of graph neural networks. Graph neural networks can effectively handle the complex relationships between tasks, represent each task and its relationship as a graph structure, and thus construct a complete tower crane task model. This model provides clear guidance for the operation of the tower crane, enabling the tower crane to perform tasks according to the predetermined steps and requirements.

[0066] Further optionally, analysis models may be pre-built for each of these operation instructions and stored in a candidate algorithm model library.

[0067] For example, the linkage table operation instructions, as structured control signals, need to be parsed into the collaborative control logic of the tower crane actuators, focusing on the timing constraints and safety thresholds of multi-mechanism actions. For this, the state machine-graph neural network maps the instructions to the state transfer of the mechanism, and uses the graph neural network to model the spatial dependencies of each mechanism, replacing the traditional rule engine, and automatically learning the linkage rules in complex scenarios. Optionally, the decision tree enhanced by the attention mechanism is used to dynamically weight the instruction timing data, generate a priority decision tree, detect abnormal operation combinations in real time, and improve operational safety and misoperation detection rate.

[0068] For example, voice command analysis needs to extract operation intentions, target objects and constraints from voice signals to achieve end-to-end processing of voice recognition and semantic analysis. Multimodal pre-training models such as UniSpeech and SpeechT5 combine voice waveforms with text semantics for pre-training, directly convert voice commands into structured operation parameters, support complex semantic analysis of long sentences, and reduce intermediate error transmission; the voice semantic alignment model enhanced by contrastive learning aligns voice features and semantic vectors through contrastive learning to solve homonymous ambiguity, improves recognition accuracy by more than 30% in noisy construction site environments, and is also compatible with dialect analysis.

[0069] For example, fixed-distance driving command parsing aims to convert movement commands into path planning and trajectory optimization while satisfying constraints such as speed and acceleration. Deep model predictive control combines deep learning with MPC, uses neural networks to model the nonlinear dynamic characteristics of tower cranes, adds safety margin constraints, dynamically adapts to environmental changes, and reduces trajectory tracking errors. Reinforcement learning-model predictive control fusion uses offline reinforcement learning pre-training strategies and MPC to generate short-term trajectories online, balancing global optimality and real-time performance to achieve rapid path adjustment in high-speed scenarios.

[0070] For example, the analysis of micro-control instructions focuses on converting high-precision operation instructions into fine control of the motor servo system, and processing high-frequency position feedback and dynamic disturbances. The neural differential equation models the motor control as a continuous-time dynamic system, and learns the differential equation parameters to achieve continuous control of current and voltage, achieving microsecond-level control accuracy and meeting high-precision docking requirements; hierarchical reinforcement learning decomposes the operation into the task layer and the control layer, hierarchical optimization strategy, reduces the state space dimension, and improves the learning efficiency and control stability of complex scenarios such as multi-axis collaborative fine-tuning.

[0071] For example, the parsing of somatosensory control instructions requires identifying the operation intention from the spatiotemporal sequence of human body movements. The spatiotemporal graph neural network models the gesture skeleton points as a graph structure, extracts the spatial joint association and time series dynamics through spatiotemporal graph convolution, supports complex continuous action recognition, and is robust to occlusion and perspective changes. The video Transformer uses the self-attention mechanism to capture the action dependency between somatosensory video frames, learns the action mode end-to-end, and achieves low-latency response.

[0072] Furthermore, in terms of candidate algorithm model library construction and adaptation, the applicable scenarios are stored and labeled by instruction type, dynamic loading and online update are supported, and federated learning is used to aggregate data from multiple construction sites to optimize generalization capabilities. The model matching strategy automatically selects the optimal analysis model based on real-time weather, lighting, noise and other status information. For example, the ST-GNN model based on the depth camera is switched for somatosensory control under strong light, and the CL-SAM model with noise-resistance enhancement is selected for voice analysis in a high-noise environment, thereby realizing full-link intelligence from instruction input to precise execution, significantly improving the operational safety and efficiency of tower cranes under complex working conditions.

[0073] Further optionally, after receiving the operation instruction issued by the user to the tower crane in step S104, the operation instruction can also be parsed to obtain the operation task requirements in the operation instruction; and it is determined whether the target mode currently in which the tower crane is located meets the operation task requirements. If the target mode does not meet the operation task requirements, the target mode is reselected based on the operation instruction; the operation instruction, the real-time operation information of the tower crane, and the reselected target mode are input into the fuzzy control model to generate the device control parameters in the target mode switching process, so as to smoothly switch the tower crane from the current target mode to the reselected target mode. The device control parameters include at least: motor speed adjustment value, braking force of the brake device, and speed of the variable amplitude mechanism.

[0074] In the above steps, after receiving the user's operation instructions, the core task requirements in the instructions are first extracted through natural language processing, signal analysis and other technologies, such as lifting weight, target position, accuracy requirements, time limit, etc. At the same time, combined with the current target mode of the tower crane equipment (such as linkage table mode, fixed-distance driving mode, etc.) and its inherent capabilities (such as fixed-distance mode is good at precise movement, linkage table mode is highly flexible but has low accuracy), a comparative analysis is performed to see whether the target mode can meet the task requirements. For example, if the task requires placing a heavy object at a specific location with millimeter-level accuracy, and the current target mode is linkage table mode, since its accuracy is difficult to meet the requirements, it is determined that the current mode does not meet the task requirements.

[0075] When it is determined that the current target mode does not meet the requirements of the task, the appropriate target mode is reselected based on the key information in the operation instruction, historical data and preset mode selection rules. For example, if the current task is high-precision lifting, the system may switch the target mode from the linkage table mode to the fixed-distance driving mode or the micro-distance control mode to better match the high-precision operation requirements. Then, the operation instructions, the real-time operation information of the tower crane equipment (such as the current hook position, the boom angle, the load weight, etc.) and the reselected target mode are input into the fuzzy control model. The fuzzy control model fuzzifies the input information through pre-set fuzzy rules, and converts the precise input data into fuzzy sets (such as "fast speed" and "strong braking force" and other fuzzy concepts). Then, according to the fuzzy reasoning rules, the fuzzy set is logically inferred to obtain the fuzzy output. Finally, through the defuzzification operation, the fuzzy output is converted into precise equipment control parameters, such as the motor speed adjustment value, the braking force of the brake device, the speed of the luffing mechanism, etc. These parameters are used to control the tower crane equipment during the mode switching process to achieve a smooth and safe transition and avoid mechanical shock or loss of control of operation caused by mode switching.

[0076] For example, suppose that at a certain construction site, the tower crane is currently in the linkage table mode, and the operator issues an operation instruction "precisely lift a 5-ton weight to a designated platform 20 meters away from the current position and 15 meters high". After parsing the instruction, it is clear that the task requirement is to lift a 5-ton weight to a specific location, and there are high requirements for accuracy. The current linkage table mode is evaluated and it is found that it is difficult to meet the high-precision lifting requirements, and it is determined that the current target mode does not meet the task requirements. According to the task requirements, the system reselects the fixed-distance driving mode as the target mode because this mode can better achieve precise distance control and positioning. Then, the operation instruction, the current real-time operation information of the tower crane (such as the current position of the hook, the current angle of the crane arm, the load weight of 5 tons, etc.) and the reselected fixed-distance driving mode are input into the fuzzy control model. The model processes according to the preset fuzzy rules. For example, considering that the load is heavy and needs to be moved accurately, the equipment control parameters generated by the fuzzy control model may be: the motor speed adjustment value is set to a lower speed to ensure a smooth lifting process; the braking force of the brake device is appropriately increased to prevent slipping during the lifting process; the speed of the variable-length mechanism is also set to a lower value to ensure that the position of the crane arm can be accurately adjusted. Through these control parameters, the tower crane smoothly switches from the linkage mode to the fixed-distance driving mode and successfully completes the lifting task.

[0077] Therefore, through dynamic evaluation and target mode reselection, the tower crane can automatically adapt the operation mode according to the operation requirements, improve the adaptability to diversified scenarios, and avoid task obstruction and efficiency loss. The parameters generated by the fuzzy control model ensure smooth mode switching, reduce mechanical impact, enhance operational stability, and extend equipment life. Accurate mode selection and stable switching reduce safety risks and prevent accidents such as overloading and collision. Automatically matching the optimal mode and fast and smooth switching reduce manual intervention, improve tower crane operation efficiency, speed up construction progress, and reduce costs.

[0078] As an optional embodiment, it is assumed that the real-time operation information includes at least: the boom position of the tower crane equipment, the hook position, the equipment movement state information, the load information, and the environmental information.

[0079] Based on the above assumptions, in the above steps, the operation instructions, the real-time operation information of the tower crane equipment, and the re-selected target mode are input into the fuzzy control model to generate the equipment control parameters in the target mode switching process, so as to smoothly switch the tower crane equipment from the current target mode to the re-selected target mode, including: taking the operation target, operation type, real-time operation information, and re-selected target mode in the operation instructions as input variables, identifying the value range and actual physical meaning of each input variable, and configuring a corresponding first fuzzy set and membership function for each input variable based on the identification result; calculating the membership of each input variable in the corresponding first fuzzy set through the membership function; performing fuzzy reasoning on the target mode switching process of the tower crane equipment based on the membership and pre-set fuzzy rules to obtain a second fuzzy set for constructing the equipment control parameters; performing defuzzification processing based on the second fuzzy set to obtain the equipment control parameters.

[0080] Specifically, in the above steps, the operation objectives (such as lifting to a specific location, placement accuracy requirements, etc.), operation types (such as ordinary lifting, precision installation, etc.) in the operation instructions, as well as real-time operation information (boom position, hook position, equipment motion state information, load information, environmental information) and the re-selected target mode are used as input variables. First, identify the value range and actual physical meaning of each input variable. Then, according to these identification results, configure the corresponding first fuzzy set and membership function (such as triangular membership function, trapezoidal membership function, etc.) for each input variable. The membership function is used to describe the degree to which the input variable belongs to a fuzzy set. Then, the membership of each input variable in the corresponding first fuzzy set is calculated by the membership function. For example, if the boom angle is 60 degrees, its membership in the fuzzy set is calculated according to the membership function of "small boom angle", and so on, the membership of all input variables in their respective fuzzy sets is calculated. Based on the calculated membership and pre-set fuzzy rules (such as "If the boom angle is large and the load is heavy, the motor speed adjustment value should be small"), fuzzy reasoning is performed on the target mode switching process of the tower crane equipment. Through these rules, the membership of the input variables is logically operated to obtain the second fuzzy set used to construct the equipment control parameters (such as "the motor speed adjustment value is small" and "the braking force of the brake device is large"). Finally, the obtained second fuzzy set is defuzzified and the fuzzy result is converted into an accurate value, that is, the equipment control parameters (such as the specific value of the motor speed adjustment value, the specific size of the braking force of the brake device, the specific value of the speed of the variable amplitude mechanism, etc.) are obtained.

[0081] For example, assuming that the tower crane is currently in linkage mode, the operator issues an operation instruction "precisely lift 8 tons of weight to a designated platform 30 meters horizontally and 20 meters vertically from the current position". After evaluation, the system reselects the fixed-distance driving mode as the target mode. Assume that the operation target is to lift to 30 meters horizontally and 20 meters vertically, with high accuracy requirements. Assume that the operation type is precision lifting. Assume that the real-time operation information is: the boom angle is 45 degrees (value range 0 to 180 degrees), configure the fuzzy set "small boom angle", "medium boom angle", "large boom angle"; the hook position is 10 meters horizontally and 8 meters vertically from the target position, configure the corresponding fuzzy set; the equipment motion state information such as the hook lifting speed is 0.5 meters / second (value range 0 to 2 meters / second), configure the fuzzy sets such as "slow speed", "medium speed", "fast speed"; the load information is the load weight of 8 tons (exceeding the medium load under normal circumstances), configure the fuzzy sets of "light load", "medium load", and "heavy load"; environmental information such as wind force is level 4 (value range 0 to 12), configure the fuzzy sets of "small wind force", "medium wind force", and "strong wind force". Assume that the target mode is fixed-distance driving mode. Based on the above assumptions, in this example, the corresponding membership function is configured for each fuzzy set, such as the "medium boom angle" fuzzy set of the boom angle, using a triangular membership function. Through the membership function calculation, for example, the membership of the boom angle of 45 degrees in the "medium boom angle" fuzzy set is 0.6; the membership of the hook position in the corresponding fuzzy set; the membership of the load weight of 8 tons in the "heavy load" fuzzy set, etc. According to the pre-set fuzzy rules, such as "if the load is heavy and the boom angle is medium, and the wind force is medium, then the motor speed adjustment value should be small and the braking force of the brake device should be large", combined with the calculated membership, reasoning is performed to obtain the second fuzzy set used to construct the equipment control parameters, such as "small motor speed adjustment value" and "large braking force of the brake device". The second fuzzy set is defuzzified using the center of gravity method to obtain equipment control parameters such as the motor speed adjustment value of 0.2 m / s, the braking force of the brake device is a specific value, and the speed of the variable amplitude mechanism is a specific value, thereby achieving a smooth switch from the linkage mode to the fixed distance driving mode.

[0082] Therefore, the fuzzy control model comprehensively processes multiple input variables such as operation objectives, types, real-time information and target modes, generates equipment control parameters, ensures smooth transition of the tower crane when switching modes, reduces mechanical shock and operation fluctuations, and improves stability and reliability. It can flexibly adjust parameters according to different working conditions and operation requirements, enhance the adaptability of the tower crane to complex environments, and improve operation safety and efficiency. In precision operations, parameters can be accurately adjusted to meet high-precision requirements. In addition, the model simulates expert decision-making based on preset fuzzy rules, provides intelligent support for tower crane operations, and reduces human intervention and decision-making errors.

[0083] Further optionally, in the above steps, defuzzification processing is performed based on the second fuzzy set to obtain the device control parameters, including: using the centroid method to determine the centroid position of the second fuzzy set according to the membership and value range of each fuzzy element in the second fuzzy set, and using the numerical value corresponding to the centroid position as the first device control parameter. Using the maximum membership method, according to the membership and value range of each fuzzy element in the second fuzzy set, the fuzzy element with the largest membership is selected, and the midpoint of the value range of the selected fuzzy element is used as the second device control parameter. Based on the data value characteristics of the device control parameters, the device control parameters for the final output are selected from the first device control parameters and the second device control parameters. The specific implementation process is explained below using the combined use of the two methods as an example.

[0084] The centroid method is a defuzzification method that determines the result by calculating the weighted average of all elements in the fuzzy set (the weight is the element membership). This method comprehensively considers the information of all elements in the set, can more comprehensively reflect the overall characteristics, is more applicable in scenarios with high control accuracy requirements, can accurately determine the equipment control parameters, and reduce errors. It is applicable to three types of scenarios: First, when it is necessary to comprehensively consider the influence of multiple factors, such as the position of the hook, speed, load and other factors in the complex operation of the tower crane, which jointly act on the adjustment of the lifting speed, and can fully utilize the fuzzy state contribution of each factor. Second, for scenarios with strict control accuracy requirements, such as the installation of components in large-scale construction projects, it can ensure accurate and safe operation. For example, in large-scale construction projects, the accuracy of the component installation position is extremely high. The centroid method can better meet the control accuracy requirements and ensure the accuracy and safety of the tower crane operation. Third, in the face of uniform or complex distribution of the membership function, when the fuzzy set has no obvious dominant element, a reasonable result can be obtained based on the actual distribution, such as the complex speed adjustment value distribution of the tower crane in different operation stages. For example, in different operation stages of a tower crane, the membership of each fuzzy state in the fuzzy set of the lifting speed adjustment value may present a complex distribution due to the influence of multiple factors. At this time, the centroid method can calculate the appropriate adjustment value based on the actual distribution.

[0085] In an example of the calculation process of the centroid method, it is assumed that there are three fuzzy elements in the second fuzzy set regarding the motor speed adjustment value, namely "increase", "unchanged" and "decrease", and their membership and value ranges are as follows: Increase: The membership is 0.3, the value range is [5%, 10%], and the representative value in this range can be set to 7.5% (usually the midpoint). Unchanged: The membership is 0.4, the value range is [-1%, 1%], and the representative value is 0%. Decrease: The membership is 0.3, the value range is [-10%, -5%], and the representative value is -7.5%.

[0086] Since the centroid method is a defuzzification method in fuzzy control, its principle is based on the idea of ​​moment balance, and the centroid position of the fuzzy set is used as the precise value after defuzzification. Suppose in a discrete fuzzy set, there is Fuzzy elements , and its corresponding membership degree is From a physical point of view, each fuzzy element It can be regarded as a quality A particle at position Then, the center of gravity of the entire fuzzy set is equivalent to the point of action of the combined force of these particles. According to the principle of moment balance, assuming that the center of gravity is The calculation formula can be expressed as: ;in, is the value corresponding to the center of gravity position, It is The representative value of the element, It is The membership degree of an element, is the number of fuzzy elements in the fuzzy set. Based on this formula, the center of gravity position (i.e., the motor speed adjustment value) corresponding to the second fuzzy set of the motor speed adjustment value in the above example is, .

[0087] Further, although the traditional centroid method works well for symmetric membership functions (such as triangles and trapezoids), it is easily affected by extreme values ​​when the membership function is asymmetric or multi-peaked. In view of this situation, the above formula can be further improved, such as introducing a weight adjustment factor, which can be set according to the physical meaning of the fuzzy element or the system security level. Based on the above centroid position The calculation formula of the improved center of gravity position The calculation formula is: ;in, is the weight adjustment factor, and the other parameters have the same meaning and will not be described here. For example, in the tower crane motor speed adjustment, the "large acceleration" (high risk state) is set is 0.5, set the "Small Adjustment" The weight of dangerous operations is reduced by 1. In this way, the influence of unreasonable extreme values ​​can be suppressed through the weight adjustment factor, thus improving control safety.

[0088] Further, in order to adapt to the control requirements of different control stages of the tower crane, the system introduces a phased dynamic configuration mechanism of the weight adjustment factor. When the tower crane switches from the "heavy load lifting stage" to the "precision positioning stage", the mechanism can intelligently adjust the control strategy: the lifting stage requires fast movement, the system increases the weight of the "speed adjustment" related fuzzy elements (such as setting it to 1.2), and reduces the "position accuracy" weight (such as setting it to 0.8), so that the center of gravity calculation is biased towards efficient movement; the positioning stage requires high-precision fine-tuning, then the weight is adjusted in the opposite direction, the "speed adjustment" weight is reduced to 0.6, and the "position deviation" related element weight is increased to 1.5, so that the calculation result meets the requirements of fine control. The system judges the stage transition by real-time monitoring of the hook height, load shaking and other states, and dynamically adjusts the influence of each fuzzy element on the center of gravity calculation, avoiding the imbalance of control of traditional fixed weights at different stages, effectively improving the lifting efficiency by 20%, reducing the positioning error by 30%, and solving the problem that a single weight is difficult to adapt to the complex control requirements of multiple stages.

[0089] The maximum membership method is to find the element with the maximum membership in the fuzzy set, and use its corresponding precise value as the defuzzification result. If there are multiple maximum membership elements, the average, minimum or maximum value can be taken. This method is simple to calculate and the result is fast. It is very suitable for tower crane control systems with high real-time requirements, and can give control parameters in time according to the fuzzy reasoning results. For example, when the tower crane urgently avoids obstacles or adjusts its position quickly, the parameters can be determined quickly; when the fuzzy reasoning has a dominant element (the membership of a certain state is much higher than that of others), the precise value of the element can be used directly to determine the parameters as the result. In scenarios where the results require high intuitiveness and simplicity, such as operator training or simple operations, directly selecting representative elements facilitates understanding and execution of parameters.

[0090] In an example of the calculation process of the maximum membership method, assuming that the fuzzy elements about the braking force of the brake device in the second fuzzy set are "strong", "medium" and "weak", their membership and value ranges are as follows: Strong: The membership is 0.2, the value range is [80,100] (assuming a certain quantitative unit of braking force), and the midpoint value is 90. Medium: The membership is 0.5, the value range is [50,70], and the midpoint value is 60. Weak: The membership is 0.3, the value range is [0,30], and the midpoint value is 15. Here, the membership of "medium" is the largest, so according to the maximum membership method, the midpoint of the value range of the fuzzy element "medium" is selected as the adjustment value of the braking force of the brake device. In this way, the equipment control parameters can be quickly determined according to the fuzzy reasoning results. In scenarios with high real-time requirements, the tower crane equipment can respond to operation instructions in a timely manner, quickly switch modes and control equipment, and improve work efficiency.

[0091] The above embodiment combines the center of gravity method with the maximum membership method to determine the tower crane control parameters. The center of gravity method calculates the weighted average of the fuzzy set elements and outputs precise parameters; the maximum membership method takes the midpoint of the range corresponding to the maximum membership element, and the result is intuitive and the calculation is fast. According to the characteristics of parameter data, the two have complementary advantages: the center of gravity method is used to ensure precise control for high-precision operations, such as hoisting precision components; the maximum membership method is used to improve the response speed of conventional fast switching scenarios, taking into account both accuracy and real-time performance, effectively enhancing the tower crane control performance, and ensuring operation safety and efficiency.

[0092] Step S105: display the tower crane task model to the user.

[0093] Step S106: After the user confirms the tower crane task model, the tower crane task model is sent to the control system to control the tower crane equipment to complete the corresponding task according to the operation sequence.

[0094] In the tower crane assisted driving system, in step S105, the task model is converted into a visual interface. The operation nodes (such as "lifting preparation → load detection → horizontal movement → precise placement → return to stop") are displayed in a timeline or flowchart, and each node is marked with the execution order, estimated time consumption, and dependency (such as "horizontal movement" must be triggered after "lifting height meets the standard"), and the real-time equipment status (such as crane arm angle, hook position) is dynamically associated. If the detection data exceeds the safety threshold, a red warning is marked. Based on the digital twin model of the construction site that integrates BIM coordinates and environmental data, the lifting path and spatial constraints are visualized, and the details can be zoomed and rotated. For example, the planned trajectory is highlighted in blue, obstacles (such as adjacent tower cranes and buildings) are marked with red translucent models, and the safety distance (such as no entry within 2 meters) is dynamically rendered in a yellow warning area. The path details can be viewed by zooming and rotating. For example, in the "precise placement" node, the millimeter-level accuracy requirements of the target position and the corresponding camera visual guidance range can be seen after zooming in. In addition, it also supports users to manually adjust task parameters or insert temporary nodes, verify the feasibility of the adjustments in real time, and update the risk assessment results.

[0095] Taking the virtual reality (VR) device as an example, it is suitable for complex working conditions (such as high-altitude heavy-load lifting). Through the immersive 3D scene, the operator feels as if he is in the tower crane working space. He can view the node details through gesture interaction (such as grabbing the "horizontal movement" node to display the angular velocity limit of the crane arm at this stage of 0.5° / s), or by turning the head to perceive the spatial relationship between the hook and the target position in real time.

[0096] As an optional embodiment, in step S105, the tower crane task model is constructed as a visual model; the visual model is displayed through a linkage platform or a virtual reality device. In the embodiment of the present application, multiple task nodes associated with the task to be executed are displayed in sequence from first to last in the visual model according to the operation sequence. In step S105 of the tower crane assisted driving system, the operation process is transparent and operable by constructing a visual task model and interactively displaying it with the help of a linkage platform or a VR device. The model construction adopts task node time-series modeling, disassembles the tower crane operation into multi-level nodes, encapsulates basic attributes, execution parameters and safety information, and ensures that the operation logic and safety strategy are fully presented; at the same time, data-driven rendering technology is used to synchronize the real-time status collected by the sensor, compare the display plan with the actual trajectory, and build a 3D scene based on BIM or CAD to mark the path and obstacles. In terms of display adaptation, the linkage platform displays the task nodes with a 2D structured interface of an industrial touch screen, and uses a Gantt chart to present the risk level and planned time. It supports clicking to view parameters, dragging to adjust the task order, and automatically checking the feasibility. VR equipment brings an immersive experience, arranging task nodes with suspended cubes, superimposing virtual models with real scenes, and supporting gesture operations to view details, making it easier for operators to perceive the deviation between the hook and the target position.

[0097] For example, taking the lifting of bridge components at a high-speed railway pier construction site, the operator views the task model through VR equipment. The "ground detection" node displays the wire rope wear and load weight in real time. The "heavy load lifting" node highlights the lifting path that avoids ground vehicles and automatically switches the control mode. The "high-altitude precision docking" node is embedded in the visual guidance interface to provide real-time feedback on the hook offset. The linkage platform simultaneously displays the Gantt chart, with the wind speed and load sway safety thresholds of the "high-altitude precision docking" node highlighted, and real-time prompts for parameter adjustment of the wind-resistant mode when the wind speed changes.

[0098] Through model-driven display and multimodal interaction, abstract task planning is transformed into a perceptible and interventional intuitive interface. The time-series node display enables operators to clearly understand the operation process and key parameters, reducing errors caused by information asymmetry.

[0099] Further optionally, it is assumed that each task node is provided with a control for managing or viewing the work tasks to be performed. Based on the above assumption, after the tower crane task model is displayed to the user in step S105, it also includes: modifying the steps, sequence, and task type in the work tasks to be performed corresponding to the task node by modifying the control; deleting the work tasks to be performed corresponding to the task node by deleting the control; adjusting the equipment control parameters in the work tasks to be performed corresponding to the task node by adjusting the parameter control; rebuilding the tower crane task model based on the tower crane task model after maintenance; generating a visualization model corresponding to the tower crane task model, and pushing it to the user end to confirm whether to execute the rebuilt tower crane task model.

[0100] Specifically, in the tower crane assisted driving system, when each task node is provided with a control for managing or viewing the work tasks to be performed, after the user views the visual task model in step S105, the task node can be dynamically maintained through the interactive control.

[0101] For example, in the 2D interface of the linkage platform or the 3D scene of the VR device, there are gear-shaped modification controls, trash can-shaped deletion controls, and slider-shaped parameter adjustment controls next to each task node: Click the modification control to pop up a hierarchical menu, allowing users to drag and adjust task steps (such as advancing the "load detection" node to before the "lifting preparation"), change the task order (such as splitting the "horizontal movement" node into two sub-nodes of "left movement to avoid obstacles → right movement to position" due to obstacles) or modify the task type (such as adjusting "normal lifting" to "precision positioning" and activating the corresponding micro-distance control strategy). Click the deletion control to remove redundant nodes (such as deleting the preset "wind resistance mode verification" node when the weather is fine), and automatically verify the logical integrity of the deleted tasks (such as confirming whether the remaining nodes meet the load safety threshold). The sliding parameter adjustment control can modify the device control parameters in real time (such as fine-tuning the motor speed of the "Precise Placement" node from 0.1m / s to 0.08m / s, or relaxing the allowable range of position deviation to ±1.5cm). During adjustment, the interface synchronously displays the impact of parameter changes on task duration and safety margin (for example, a reduction in speed increases the duration by 2 minutes and increases the safety margin by 15%).

[0102] After the user completes the maintenance operation, the matching relationship between the operation instructions and the real-time operation information is re-analyzed based on the updated task steps, sequence, type and parameters. For example, when the task type is changed from "normal lifting" to "precision positioning", the higher-precision visual servo algorithm and PID control parameters are automatically loaded, and the multimodal data fusion verification is triggered (such as replanning the path by combining the hook position sensor and the visual camera data). The reconstructed tower crane task model will generate a new visual interface, highlight the changed part in the form of a Gantt chart on the linkage platform (such as the adjusted node is marked with an orange border), and prompt the change of the task sequence with a flashing cube in the VR device. At the same time, a confirmation pop-up window containing the modification summary (such as "adding an obstacle avoidance node, the estimated time is increased by 5 minutes"), parameter adjustment details (such as "the motor speed is reduced by 20%, and the accuracy is improved to ±1cm") and risk assessment results (such as "the collision risk is reduced from 35% to 12%") is pushed for the user to confirm again. If the user confirms the execution, the control instruction flow will be generated based on the maintained model. If there is a logical conflict (such as deleting a key safety verification node), a pop-up window will be displayed to mark the cause of the error (such as "load detection node is missing, overload risk cannot be verified") and the modification will be locked to ensure the safety and feasibility of the task model. This dynamic maintenance mechanism with control interaction enables operators to flexibly adjust the operation plan according to real-time working conditions, such as quickly inserting obstacle avoidance steps when new obstacles are found, or temporarily modifying control parameters when the equipment status is abnormal. A closed loop is formed through real-time verification and visual feedback, which not only retains the efficiency of intelligent planning, but also gives flexibility to manual intervention, significantly improving the task adaptability and operational safety of tower cranes in complex construction sites.

[0103] In step S106, after the user confirms the tower crane task model, a two-way check is first performed to ensure the reliability and safety of the task. On the one hand, the task node timing logic is verified to ensure that the action nodes such as "descending" are triggered after the preconditions such as "reaching above the target position" are met, and at the same time, the matching of the control strategy and the equipment capabilities is checked (such as the "macro control" node is only activated when the hook speed is <0.1m / s). On the other hand, based on the tower crane dynamics model and real-time environmental parameters (such as wind speed, load weight), the safety margin of each node is calculated (such as the ratio of load weight to rated load, the difference between obstacle distance and minimum safety distance). If the check fails (such as margin <20%), the user interface is automatically returned and the risk node (such as the "horizontal movement" node of "heavy load + close obstacle") is highlighted to prompt the user to correct it.

[0104] Further optionally, in step S106 of the tower crane auxiliary driving system, the verified task model is converted into a control system instruction stream through the industrial protocol adaptation layer. Using graph structure coding technology, the task node is disassembled into hardware control signals and software strategy calls under the binary protocol. For example, the task model of the graph structure (including nodes and edge relationships) is encoded into binary industrial protocols such as Profinet and EtherCAT, and each task node is disassembled into the underlying control signal. For example, "lifting" corresponds to the motor operation instruction, and "precise placement" triggers the visual servo module. At the same time, a priority control sequence is generated according to the risk level and time constraints, interrupt authority is granted to the safety instruction, and the instruction is encrypted and transmitted through digital signatures to ensure the safety and reliability of the instruction. After receiving the instruction, intelligent execution is achieved with the help of the task scheduler: the corresponding control module is activated according to the job sequence, the progress is dynamically monitored in combination with the sensor data, and the task rhythm is automatically adjusted when the environment changes. For example, the path planning algorithm is called when executing the "horizontal movement" node, the pre-trained dynamic compensation parameters are loaded when executing the "wind resistance mode" node, and the execution progress is dynamically monitored through the data such as the hook position and the boom angle transmitted back by the sensor in real time. For example, if there are environmental changes such as a sudden increase in wind speed, the "horizontal movement" speed will be automatically shortened and the subsequent node time window will be extended to ensure adaptive adjustment of the task rhythm. When risks such as obstacle intrusion and load shaking exceeding the threshold are detected, the "safe fallback" mechanism will be immediately triggered to suspend the operation, switch to safe mode, and push abnormal details. After the risk is eliminated, it will be restored in an orderly manner to form a closed-loop control of "planning-execution-feedback-adjustment". Furthermore, through the visual interactive interface, the operator's task confirmation time is shortened and dynamic adjustment is supported. The two-way verification mechanism can reduce the logical error rate and avoid most safety risks. Real-time synchronization of plans and actual status can be applied to complex working conditions such as "heavy load + strong wind", improve the success rate of operations, and realize the intelligence, efficiency and safety of tower crane operations.

[0105] In the embodiment of the present application, the intelligent control of the tower crane equipment is realized through the adaptive decision model, real-time mode configuration and tower crane task model, which helps to improve the adaptability of operators to complex working conditions, reduce human errors and ensure the safety of operations. At the same time, with the help of a variety of assisted driving modes, the difficulty of operation can be reduced and the efficiency of equipment control can be improved. In particular, the adaptive decision model has the ability of autonomous learning, and can improve the control efficiency of the tower crane under complex working conditions through the integration of a variety of assisted driving modes and intelligent decision-making mechanisms, thereby comprehensively improving the efficiency, accuracy and safety of operations.

[0106] See also Figure 2 , Figure 2A tower crane control system 200 based on assisted driving is provided in the embodiment of the present application. The tower crane control system 200 based on assisted driving includes a deployment module 201, a decision module 202, a configuration module 203, a task module 204, a display module 205, and a sending module 206. The deployment module 201 is used to deploy the control system in the tower crane equipment; the control system is connected to the linkage platform, and a plurality of assisted driving modes are integrated in the control system; the plurality of assisted driving modes are respectively used to assist the user to adopt the corresponding operation mode to realize the control operation of the tower crane equipment; the plurality of assisted driving modes include at least: linkage platform mode, voice driving mode, fixed distance driving mode, micro distance control mode, and body sensing control mode; the decision module 202 is used to construct an adaptive decision model of the tower crane equipment using the A3C algorithm according to the historical status information of the tower crane equipment, the historical mode selection information, and the historical execution status of the tower crane operation task, so as to select the assisted driving mode that best matches the tower crane equipment; the configuration module 203 is used to obtain the real-time status information of the tower crane equipment, and adopt the adaptive decision model The target mode that matches the real-time status information is configured; the task module 204 is used to receive the operation instructions issued by the user to the tower crane equipment, and to construct the tower crane task model of the tower crane equipment in the target mode based on the operation instructions; the tower crane task model is provided with multiple task nodes in order from first to last according to the operation sequence, and each task node corresponds to a task to be executed; the operation instruction is at least one of the linkage table operation instruction, voice instruction, fixed-distance driving instruction, micro-distance control instruction, and somatosensory control instruction; the display module 205 is used to display the tower crane task model to the user; the sending module 206 is used to send the tower crane task model to the control system after the user confirms the tower crane task model, so as to control the tower crane equipment to complete the corresponding task according to the operation sequence. In some embodiments, the tower crane control system 200 based on assisted driving can be applied to the terminal device. It should be noted that, for the convenience and simplicity of description, the specific working process of the tower crane control system 200 based on assisted driving described above can refer to the corresponding process in the embodiment of the tower crane control method based on assisted driving, which will not be repeated here.

[0107] See also Figure 3 , Figure 3 A schematic block diagram of the structure of a terminal device provided in an embodiment of the present application.

[0108] like Figure 3As shown, the terminal device 300 includes a processor 301 and a memory 302, and the processor 301 and the memory 302 are connected via a bus 303, such as an I2C bus. Specifically, the processor 301 is used to provide computing and control capabilities to support the operation of the entire terminal device. The processor 301 can be a central processing unit, and the processor 301 can also be other general-purpose processors, digital signal processors, application-specific integrated circuits, field programmable gate arrays or other programmable logic devices, discrete gates or transistor logic devices, discrete hardware components, etc. Among them, the general-purpose processor can be a microprocessor or the processor can also be any conventional processor, etc. Specifically, the memory 302 can be a Flash chip, a read-only memory disk, an optical disk, a USB flash drive or a mobile hard disk, etc.

[0109] Those skilled in the art will understand that Figure 3 The structure shown in is only a block diagram of a part of the structure related to the embodiment of the present application, and does not constitute a limitation on the terminal device to which the embodiment of the present application is applied. The specific server may include more or fewer components than shown in the figure, or combine certain components, or have a different arrangement of components. Among them, the processor is used to run a computer program stored in the memory, and implement any one of the tower crane control methods based on assisted driving provided in the embodiment of the present application when executing the computer program. It should be noted that technical personnel in the relevant field can clearly understand that for the convenience and conciseness of description, the specific working process of the terminal device described above can refer to the aforementioned embodiment of the tower crane control method based on assisted driving, and will not be repeated here.

[0110] An embodiment of the present application also provides a storage medium for computer-readable storage, wherein the storage medium stores one or more programs, and the one or more programs can be executed by one or more processors to implement any step of the tower crane control method based on assisted driving as provided in the description of the embodiment of the present application.

Claims

1. A tower crane control method based on assisted driving, characterized in that: The method comprises: A control system is deployed in the tower crane equipment; the control system is connected to the linkage platform, and a plurality of assisted driving modes are integrated in the control system; the plurality of assisted driving modes are respectively used to assist the user to adopt corresponding operation modes to realize the control operation of the tower crane equipment; the plurality of assisted driving modes at least include: linkage platform mode, voice driving mode, fixed distance driving mode, micro distance control mode, and somatosensory control mode; According to the historical status information, historical mode selection information and historical execution status of the tower crane operation tasks, the A3C algorithm is used to build an adaptive decision model for the tower crane to select the assisted driving mode that best matches the tower crane. Acquire real-time status information of the tower crane equipment, and use the adaptive decision model to configure a target mode that matches the real-time status information; Receive an operation instruction issued by a user to a tower crane device, and construct a tower crane task model of the tower crane device in a target mode based on the operation instruction; a plurality of task nodes are arranged in the tower crane task model in order from the first to the last according to the operation sequence, and each task node corresponds to a to-be-executed operation task; the operation instruction is at least one of a linkage platform operation instruction, a voice instruction, a fixed-distance driving instruction, a micro-distance control instruction, and a somatosensory control instruction; The tower crane task model is displayed to the user; after the user confirms the tower crane task model, the tower crane task model is sent to the control system to control the tower crane equipment to complete the corresponding task according to the operation sequence.

2. The method according to claim 1, characterized in that After receiving the operation instruction issued by the user to the tower crane equipment, the method further includes: Parsing the operation instruction to obtain the operation task requirements in the operation instruction; Determine whether the current target mode of the tower crane equipment meets the requirements of the operation task; If the target mode does not meet the task requirements, reselect the target mode based on the operation instruction; The operation instructions, the real-time operation information of the tower crane equipment, and the re-selected target mode are input into the fuzzy control model to generate the equipment control parameters in the target mode switching process, so as to smoothly switch the tower crane equipment from the current target mode to the re-selected target mode; the equipment control parameters include at least: the motor speed adjustment value, the braking force of the braking device, and the speed of the luffing mechanism.

3. The method according to claim 2, characterized in that The real-time operation information includes at least: the boom position, hook position, equipment motion status information, load information, and environmental information of the tower crane; The operation instructions, the real-time operation information of the tower crane equipment, and the re-selected target mode are input into the fuzzy control model to generate the equipment control parameters in the target mode switching process, so as to smoothly switch the tower crane equipment from the current target mode to the re-selected target mode, including: Taking the operation target, the operation type, the real-time operation information, and the re-selected target mode in the operation instruction as input variables, identifying the value range and actual physical meaning of each input variable, and configuring a corresponding first fuzzy set and a membership function for each input variable based on the identification result; Calculate the membership degree of each input variable in the first fuzzy set corresponding to each input variable through the membership function; Based on the membership degree and the preset fuzzy rules, fuzzy reasoning is performed on the target mode switching process of the tower crane equipment to obtain a second fuzzy set for constructing the equipment control parameters; Defuzzification processing is performed based on the second fuzzy set to obtain the device control parameter.

4. The method according to claim 3, characterized in that The defuzzification process is performed based on the second fuzzy set to obtain the device control parameter, including: Using a gravity center method, according to the membership degree and value range of each fuzzy element in the second fuzzy set, determine the gravity center position of the second fuzzy set, and use the value corresponding to the gravity center position as the first device control parameter; Using the maximum membership method, according to the membership and value range of each fuzzy element in the second fuzzy set, select the fuzzy element with the largest membership, and use the midpoint of the value range of the selected fuzzy element as the second device control parameter; Based on the data value characteristics of the device control parameter, the device control parameter that is finally output is selected from the first device control parameter and the second device control parameter.

5. The method according to claim 1, characterized in that According to the historical status information, historical mode selection information, and historical execution status of the tower crane operation tasks, the A3C algorithm is used to construct an adaptive decision model for the tower crane to select the assisted driving mode that best matches the tower crane, including: A plurality of assisted driving modes are defined as an action space, and an assisted driving mode is selected from the action space at each time step; A reward function of the adaptive decision model is set based on historical execution conditions; the reward function includes at least: task completion reward, completion time reward, completion quality reward, and abnormal situation penalty for the tower crane equipment; Obtain the historical status information of the tower crane equipment, historical mode selection information, and historical execution status of the tower crane operation tasks; Build multiple job task environments based on historical status information; Input historical status information and historical mode selection information into multiple task environments, and train the Actor network used to select the assisted driving mode in parallel; Input historical status information, historical mode selection information, and historical execution status into multiple job task environments and use them in parallel to evaluate the value of the Actor network in the Critic network; The reward function is used to update the network parameters of the Actor network and the Critic network to obtain an adaptive decision model of the tower crane equipment.

6. The method according to claim 5, characterized in that The adopting the adaptive decision model to configure a target mode matching the real-time status information includes: Acquire image data around the tower crane device, and locate obstacles based on the image data to obtain obstacle location data of the tower crane device; Analyze the real-time status information to obtain the real-time equipment component status, real-time weather conditions, and real-time wind speed conditions of the tower crane equipment; The real-time equipment component status, the real-time weather conditions, the real-time wind speed conditions, and the obstacle positioning data are input into the adaptive decision model, and the Actor network is used to predict the driving mode to obtain the prediction probability of the tower crane equipment in each auxiliary driving mode; based on the prediction probability, a candidate mode is randomly selected by roulette; Monitor the safety risk of the tower crane equipment in the candidate mode; if the safety risk is greater than a set threshold, set the target mode to an assisted driving mode that matches the safety risk item, or switch the target mode to an assisted driving mode with a higher safety level than the current target mode.

7. The method according to claim 1, characterized in that The step of constructing a tower crane task model of the tower crane equipment in the target mode based on the operation instruction includes: Selecting a target parsing model that matches the operation instruction from a candidate algorithm model library; The operation instruction is processed by using a target parsing model to obtain an operation instruction feature vector; the operation instruction feature vector includes the instruction sequence between various operation tasks, the task type to which the operation task belongs, and the logical relationship; Determine the operation tasks to be executed and the arrangement of the operation tasks based on the operation instruction feature vector and the association relationship between the preset task type and the task priority; According to the work tasks to be performed and the arrangement of the work tasks, the tower crane task model is obtained by combining graph neural networks.

8. The method according to claim 1, characterized in that The presenting the tower crane task model to the user comprises: Constructing the tower crane task model into a visual model; The visualization model is displayed through a linkage table or a virtual reality device; the visualization model displays multiple task nodes associated with the task to be executed in order from the beginning to the end according to the job sequence; Each task node is provided with a control for managing or viewing the task to be performed; after displaying the tower crane task model to the user, it also includes: By modifying the controls, modify the steps, sequence, and task type in the task to be executed corresponding to the task node; Delete the pending job tasks corresponding to the task node by deleting the control; Adjust the device control parameters in the task to be executed corresponding to the task node through the parameter adjustment control; The tower crane task model is reconstructed based on the maintained tower crane task model; a visualization model corresponding to the tower crane task model is generated and pushed to the user end to confirm whether to execute the reconstructed tower crane task model.

9. A tower crane control system based on assisted driving, characterized in that: The system comprises: A deployment module is used to deploy a control system in a tower crane; the control system is connected to a linkage platform, and a plurality of assisted driving modes are integrated in the control system; the plurality of assisted driving modes are respectively used to assist a user to adopt a corresponding operation mode to realize control operation of the tower crane; the plurality of assisted driving modes at least include: a linkage platform mode, a voice driving mode, a fixed distance driving mode, a micro distance control mode, and a somatosensory control mode; The decision-making module is used to construct an adaptive decision-making model of the tower crane using the A3C algorithm based on the historical status information of the tower crane, the historical mode selection information, and the historical execution status of the tower crane operation tasks, so as to select the assisted driving mode that best matches the tower crane; A configuration module, used to obtain real-time status information of the tower crane equipment, and use the adaptive decision model to configure a target mode that matches the real-time status information; The task module is used to receive the operation instructions issued by the user to the tower crane equipment, and to construct a tower crane task model of the tower crane equipment in the target mode based on the operation instructions; the tower crane task model is provided with a plurality of task nodes in sequence from the first to the last according to the operation sequence, and each task node corresponds to a work task to be performed; the operation instruction is at least one of the linkage platform operation instruction, voice instruction, fixed-distance driving instruction, micro-distance control instruction, and somatosensory control instruction; A display module, used to display the tower crane task model to the user; The sending module is used to send the tower crane task model to the control system after the user confirms the tower crane task model, so as to control the tower crane equipment to complete the corresponding operation task according to the operation sequence.

10. A terminal device, characterized in that: The terminal device includes a processor and a memory; The memory is used to store computer programs; The processor is used to execute the computer program and implement the tower crane control method based on assisted driving as described in any one of claims 1 to 8 when executing the computer program.

Images