Grab bucket anti-swing AI identification system for grab bucket ship unloader

Abstract

The invention relates to a grab bucket anti-swing AI recognition system for a grab bucket ship unloader, and the system specifically comprises the following modules: a camera module which is provided with two high-definition industrial-grade cameras, is equipped with an adjustable holder device, and is installed above a grab bucket; the image processing module is provided with a primary processing unit, an AI identification unit and a coordinate conversion unit and is used for receiving the video information from the camera module; the anti-swing control module comprises a calculation unit and an instruction output unit, and calculates and outputs an anti-swing control instruction according to the data output by the image processing module; the mechanism control module is provided with a driving motor, a frequency converter and a sensing device, and is used for receiving an instruction of the anti-swing control module and adjusting the movement of the trolley and the rope so as to stabilize the position of the grab bucket; and the main controller module adopts a high-performance industrial computer to coordinate the work of each module. The grab bucket ship unloader solves the problem that in the operation process of a traditional grab bucket ship unloader, a grab bucket is prone to shaking in the lifting and moving processes.

Description

Technical Field

[0001] The present invention relates to the field of port machinery automation, and in particular to an AI recognition system for grab bucket anti-swaying of a grab bucket ship unloader. Background Art

[0002] During the operation of traditional grab ship unloaders, the grab is prone to shaking during the lifting and moving process due to the influence of various factors such as sea breeze, waves, and ship shaking. This not only increases the difficulty of operation and reduces operation efficiency, but may also cause damage to the ship structure and cargo, and even cause safety accidents. Summary of the invention

[0003] In view of the problems in the prior art, the present invention provides an AI recognition system for grab bucket anti-sway for grab bucket ship unloader, which aims to realize accurate anti-sway control of grab bucket during grab bucket ship unloading by integrating high-definition camera, AI image processing, advanced control algorithm and remote monitoring technology, and improve operation efficiency and safety. The system specifically includes the following modules: Camera module: Two high-definition industrial-grade cameras are equipped with adjustable pan-tilt devices, which are installed above the grab bucket and cover the movement trajectory of the grab bucket. The camera has low light sensitivity, a resolution of 3840x2160 pixels, supports H.265 video encoding format, a frame rate of not less than 60fps, and a wide-angle design to ensure that the viewing angle covers at least 90 degrees. The built-in infrared LED light supports clear imaging at night and in low-light environments. This module can collect high-definition video information of the grab bucket in real time, providing a reliable data source for subsequent processing; the high-definition camera combined with the adjustable pan-tilt device ensures all-round and no-dead-angle monitoring; the low light sensitivity and infrared night vision function adapt to complex working environments; H.265 encoding improves data transmission efficiency; improves video acquisition quality, enhances the system's adaptability to environmental changes, and provides accurate data support for subsequent processing.

[0004] Image processing module: includes preliminary processing unit, AI recognition unit and coordinate conversion unit. The preliminary processing unit uses FPGA to perform preprocessing operations such as video denoising and enhancement; the AI ​​recognition unit deploys a deep learning model based on the TensorFlow framework to achieve fast and accurate recognition and posture estimation of the grab bucket; the coordinate conversion unit uses a perspective transformation algorithm to convert pixel coordinates into three-dimensional space coordinates. This module can efficiently process video data and generate real-time position and posture data of the grab bucket; FPGA accelerates preprocessing, the TensorFlow deep learning model improves recognition accuracy, and the perspective transformation algorithm achieves accurate coordinate conversion; it improves image processing speed and accuracy, providing a reliable basis for anti-sway control.

[0005] Anti-sway control module: includes a calculation unit and an instruction output unit. The calculation unit uses a high-performance DSP with a main frequency of not less than 800MHz and a floating-point operation unit, which can quickly execute complex anti-sway algorithm calculations; the instruction output unit communicates with the mechanism control module through the CAN bus, and the communication rate reaches 1Mbps. This module calculates and outputs anti-sway control instructions based on the data output by the image processing module; the high-performance DSP ensures fast calculation, and the CAN bus ensures efficient communication; it realizes real-time and accurate anti-sway control and improves operation stability.

[0006] Mechanism control module: It adopts high-torque, low-noise permanent magnet synchronous motor with a rated power of not less than 50kW and supports four-quadrant operation; the inverter adopts vector control strategy with a dynamic response time of less than 5 milliseconds; the sensor device includes a high-precision encoder and an acceleration sensor to ensure accurate control of the motor speed and position. The module receives anti-sway control instructions, adjusts the movement of the trolley and rope, and realizes stable control of the grab bucket; high-performance motors and inverters are combined with high-precision sensors to achieve precise control; improve control accuracy and response speed to ensure stable operation of the grab bucket.

[0007] Main controller module: adopts dual CPU redundant architecture design, each CPU is a multi-core processor, the main frequency is not less than 2GHz, equipped with at least 16GB RAM and 512GB SSD solid state drive; with hot-swap function. As the core of the system, this module coordinates the work of each module, receives and processes feedback information, and ensures stable operation of the system; the dual CPU redundant architecture improves system reliability, and the high-performance hardware configuration guarantees processing capabilities; enhances system stability and reliability, and improves overall performance.

[0008] Remote monitoring module: connects to the main controller module through a wireless network to achieve remote monitoring and fault diagnosis functions. Supports multiple terminal access, including computers, mobile phones and tablets. This module enables operators to remotely monitor the status of equipment, discover and handle faults in a timely manner; through wireless network connection and multi-terminal support, remote monitoring and diagnosis can be achieved; it can effectively improve equipment maintenance efficiency and reduce operating costs.

[0009] The purpose of the present invention is to provide an AI recognition system for grab bucket anti-sway for grab bucket ship unloaders, which realizes real-time monitoring and precise control of grab bucket movement by integrating high-definition cameras, advanced image processing technology, efficient anti-sway control algorithms and remote monitoring functions, effectively reduces grab bucket shaking and improves operating efficiency and safety.

[0010] The beneficial effects of the present invention are as follows: through precise anti-sway control, the operation interruption and repeated operation caused by grab bucket shaking are reduced, and the efficiency of ship unloading operation is significantly improved; the grab bucket shaking is effectively suppressed, the potential risk of damage to the ship structure and cargo is reduced, and the human operation errors caused by shaking are reduced, and the operation safety is improved; the high-definition camera, AI image processing, advanced control algorithm and remote monitoring technology are integrated to realize real-time monitoring, intelligent identification and precise control of grab bucket movement, and promote the development of port machinery and equipment towards intelligence and automation; the low light sensitivity characteristics and infrared night vision function of the camera module enable the system to work normally at night and in low light environments, and enhance the adaptability of the system to different operating environments; the main controller module adopts a dual CPU redundant architecture design and an efficient communication mechanism between modules to ensure the stability and reliability of the system under high-intensity operation and reduce the failure rate; the remote monitoring module supports multiple terminal access, so that the operator can monitor the equipment status anytime and anywhere, discover and handle faults in time, reduce on-site maintenance costs, and improve management efficiency; by precisely controlling the grab bucket movement, unnecessary energy consumption and emissions are reduced, which is in line with the development trend of green ports and is conducive to environmental protection and sustainable development. BRIEF DESCRIPTION OF THE DRAWINGS

[0011] The present invention is further described below in conjunction with the accompanying drawings and embodiments.

[0012] Figure 1 It is a system architecture diagram of the present invention.

[0013] Figure 2 It is the work flow chart of the present invention. DETAILED DESCRIPTION

[0014] In order to make the technical means, creative features, objectives and effects achieved by the present invention easy to understand, the present invention is further explained below in conjunction with specific implementation methods.

[0015] like Figure 1-Figure 2 As shown, the grab bucket anti-sway AI recognition system for grab bucket ship unloader described in the present invention specifically includes the following modules: Camera module: Two high-definition industrial-grade cameras are equipped with adjustable pan-tilt devices, which are installed above the grab bucket and cover the movement trajectory of the grab bucket. The camera has low light sensitivity, a resolution of 3840x2160 pixels, supports H.265 video encoding format, a frame rate of not less than 60fps, and a wide-angle design to ensure that the viewing angle covers at least 90 degrees. The built-in infrared LED light supports clear imaging at night and in low-light environments. This module can collect high-definition video information of the grab bucket in real time, providing a reliable data source for subsequent processing; the high-definition camera combined with the adjustable pan-tilt device ensures all-round and no-dead-angle monitoring; the low light sensitivity and infrared night vision function adapt to complex working environments; H.265 encoding improves data transmission efficiency; improves video acquisition quality, enhances the system's adaptability to environmental changes, and provides accurate data support for subsequent processing.

[0016] Image processing module: includes preliminary processing unit, AI recognition unit and coordinate conversion unit. The preliminary processing unit uses FPGA to perform preprocessing operations such as video denoising and enhancement; the AI ​​recognition unit deploys a deep learning model based on the TensorFlow framework to achieve fast and accurate recognition and posture estimation of the grab bucket; the coordinate conversion unit uses a perspective transformation algorithm to convert pixel coordinates into three-dimensional space coordinates. This module can efficiently process video data and generate real-time position and posture data of the grab bucket; FPGA accelerates preprocessing, the TensorFlow deep learning model improves recognition accuracy, and the perspective transformation algorithm achieves accurate coordinate conversion; it improves image processing speed and accuracy, providing a reliable basis for anti-sway control.

[0017] Anti-sway control module: includes a calculation unit and an instruction output unit. The calculation unit uses a high-performance DSP with a main frequency of not less than 800MHz and a floating-point operation unit, which can quickly execute complex anti-sway algorithm calculations; the instruction output unit communicates with the mechanism control module through the CAN bus, and the communication rate reaches 1Mbps. This module calculates and outputs anti-sway control instructions based on the data output by the image processing module; the high-performance DSP ensures fast calculation, and the CAN bus ensures efficient communication; it realizes real-time and accurate anti-sway control and improves operation stability.

[0018] Mechanism control module: It adopts high-torque, low-noise permanent magnet synchronous motor with a rated power of not less than 50kW and supports four-quadrant operation; the inverter adopts vector control strategy with a dynamic response time of less than 5 milliseconds; the sensor device includes a high-precision encoder and an acceleration sensor to ensure accurate control of the motor speed and position. The module receives anti-sway control instructions, adjusts the movement of the trolley and rope, and realizes stable control of the grab bucket; high-performance motors and inverters are combined with high-precision sensors to achieve precise control; improve control accuracy and response speed to ensure stable operation of the grab bucket.

[0019] Main controller module: adopts dual CPU redundant architecture design, each CPU is a multi-core processor, the main frequency is not less than 2GHz, equipped with at least 16GB RAM and 512GB SSD solid state drive; with hot-swap function. As the core of the system, this module coordinates the work of each module, receives and processes feedback information, and ensures stable operation of the system; the dual CPU redundant architecture improves system reliability, and the high-performance hardware configuration guarantees processing capabilities; enhances system stability and reliability, and improves overall performance.

[0020] Remote monitoring module: connects to the main controller module through a wireless network to achieve remote monitoring and fault diagnosis functions. Supports multiple terminal access, including computers, mobile phones and tablets. This module enables operators to remotely monitor the status of equipment, discover and handle faults in a timely manner; through wireless network connection and multi-terminal support, remote monitoring and diagnosis can be achieved; it can effectively improve equipment maintenance efficiency and reduce operating costs.

[0021] The specific implementation details of the present invention are as follows: Two high-definition industrial cameras are mounted on an adjustable pan / tilt above the grab bucket to ensure that the camera field of view covers the entire movement trajectory of the grab bucket. Camera parameters, including focal length and exposure, are adjusted to meet the shooting requirements under different lighting conditions.

[0022] An FPGA chip is configured in the preliminary processing unit, and the corresponding firmware program is written to realize pre-processing functions such as video denoising and enhancement. The AI ​​recognition unit deploys a deep learning model based on the TensorFlow framework. The model needs to be trained in advance on a large number of labeled data sets to accurately identify the position and posture of the grab. The coordinate conversion unit converts the processed image coordinates into actual coordinates in three-dimensional space according to the camera parameters and the preset perspective transformation algorithm.

[0023] In the computing unit of the anti-sway control module, an anti-sway algorithm program is written. The program can receive the position and posture data from the image processing module, and calculate the real-time additional speed instructions of the trolley and the rope in combination with the preset anti-sway control strategy. The instruction output unit quickly sends the calculated instructions to the mechanism control module through the CAN bus protocol.

[0024] The parameters of the permanent magnet synchronous motor in the mechanism control module are configured, including rated power, speed range, etc. The inverter is programmed with a vector control strategy to ensure that the speed and position of the motor can be accurately controlled. At the same time, the high-precision encoder and acceleration sensor are calibrated to ensure the accuracy of their measurement data. During the debugging process, the adjustment effect of the mechanism control module on the motor output is verified through simulated signals or actual operation tests.

[0025] Each module is connected to the main controller module through a high-speed communication interface. The main controller module adopts a dual CPU redundant architecture design to ensure that it can seamlessly switch to the backup CPU to continue working when the main CPU fails. A coordination control program is written in the main controller module, which can receive and process feedback information from each module and monitor the operating status of the entire system.

[0026] A wireless network interface is configured in the remote monitoring module to ensure a stable connection with the main controller module. Remote monitoring software is developed, which supports access from multiple terminals and provides functions such as real-time video monitoring, data query, and fault diagnosis. Through the remote monitoring software, operators can understand the operating status of the equipment anytime and anywhere and perform remote operation and maintenance.

[0027] After completing the individual debugging of each module, the system is integrated and tested. By simulating the actual operation scenario, the operation effect and stability of the entire system are tested. During the test, key indicators such as the shaking amplitude of the grab, operation efficiency, safety, and the system's fault recovery capability are focused on. The system is optimized and adjusted based on the test results to ensure optimal performance.

[0028] To sum up, the grab bucket anti-sway AI recognition system for grab bucket unloader of the present invention not only improves the efficiency and safety of port unloading operations, but also enhances the intelligence, environmental adaptability and reliability of the system, and provides strong support for the automation and intelligent upgrading of port machinery and equipment. It has significant technical innovation and broad application prospects.

[0029] It will be apparent to those skilled in the art that the invention is not limited to the details of the exemplary embodiments described above and that the invention can be implemented in other specific forms without departing from the spirit or essential features of the invention. Therefore, the embodiments should be considered exemplary and non-limiting in all respects, and the scope of the invention is defined by the appended claims rather than the foregoing description, and it is intended that all variations falling within the meaning and scope of the equivalent elements of the claims be included in the invention. Any reference numeral in a claim should not be considered as limiting the claim to which it relates.

[0030] In addition, it should be understood that although the present specification is described according to implementation modes, not every implementation mode contains only one independent technical solution. This description of the specification is only for the sake of clarity. Those skilled in the art should regard the specification as a whole. The technical solutions in each embodiment may also be appropriately combined to form other implementation modes that can be understood by those skilled in the art.

Claims

1. An AI recognition system for grab bucket anti-swaying of grab bucket ship unloader, characterized in that Specifically includes the following modules: The camera module is equipped with two high-definition industrial-grade cameras with an adjustable pan-tilt device, which is installed above the grab bucket and covers the movement trajectory of the grab bucket, and is used to collect video information of the grab bucket in real time; The image processing module is provided with a preliminary processing unit, an AI recognition unit, and a coordinate conversion unit, receives video information from the camera module, and generates real-time position and posture data of the grab bucket through AI algorithm processing; The anti-sway control module includes a calculation unit and an instruction output unit, which calculates and outputs an anti-sway control instruction according to the data output by the image processing module; The mechanism control module is provided with a drive motor, a frequency converter and a sensor device, receives the instruction of the anti-sway control module, and adjusts the movement of the trolley and the rope to stabilize the position of the grab bucket; The main controller module uses a high-performance industrial computer to coordinate the work of each module and receive and process the feedback information of each module.

2. The grab bucket anti-sway AI recognition system for grab bucket ship unloader according to claim 1, characterized in that: The two high-definition industrial-grade cameras used in the camera module have low light sensitivity characteristics, a resolution of 3840x2160 pixels, support for H.265 video encoding format, a frame rate of not less than 60fps, a wide-angle design camera viewing angle covering at least 90 degrees, and a built-in infrared LED light to support clear imaging at night and in low-light environments.

3. The grab bucket anti-sway AI recognition system for grab bucket ship unloader according to claim 1, characterized in that: The preliminary processing unit in the image processing module uses FPGA to perform preprocessing operations such as video denoising and enhancement; the AI ​​recognition unit deploys a convolutional neural network model based on the TensorFlow framework, with a model size not exceeding 500MB and a recognition speed not exceeding 50 milliseconds per frame; The coordinate conversion unit has a built-in perspective transformation algorithm, which accurately converts the pixel coordinates in the camera image into actual coordinates in three-dimensional space, with a conversion error of less than 1%.

4. The grab bucket anti-sway AI recognition system for grab bucket ship unloader according to claim 1, characterized in that: The computing unit in the anti-sway control module adopts a high-performance DSP with a main frequency of not less than 800MHz and a floating-point operation unit, which can quickly execute complex anti-sway algorithm calculations; the instruction output unit communicates with the trolley control unit and the rope control unit through the CAN bus, and the communication rate reaches 1Mbps.

5. The grab bucket anti-sway AI recognition system for grab bucket ship unloader according to claim 1, characterized in that: The mechanism control module adopts a high-torque, low-noise permanent magnet synchronous motor with a rated power of not less than 50kW and supports four-quadrant operation; the frequency converter adopts a vector control strategy with a dynamic response time of less than 5 milliseconds, which can accurately control the speed and position of the motor; the sensing device adopts a high-precision encoder combined with an acceleration sensor, the encoder resolution is not less than 17 bits, and the acceleration sensor sensitivity is better than 0.01g.

6. The grab bucket anti-sway AI recognition system for grab bucket ship unloader according to claim 1, characterized in that: The main controller module adopts a dual CPU redundant architecture design, each CPU is a multi-core processor with a main frequency of not less than 2GHz, equipped with at least 16GB RAM and 512GB SSD solid state drive; the main controller module has hot-swappable function.

7. The grab bucket anti-sway AI recognition system for grab bucket ship unloader according to claim 1, characterized in that: The system also includes a remote monitoring module, which is connected to the main controller module via a wireless network to realize remote monitoring and fault diagnosis functions. The remote monitoring module supports access from multiple terminals, including computers, mobile phones and tablets.

8. The grab bucket anti-sway AI recognition system for grab bucket ship unloader according to claim 1 further comprises an operation method, characterized in that The specific steps include: S1: The camera module continuously collects video data at a preset frame rate and transmits it to the image processing module through a high-speed Ethernet interface; S2: The image processing module processes the video data in real time, with each frame processing time not exceeding 50 milliseconds, and sends the processing results, grab position and posture data to the anti-sway control module; S3: The anti-sway control module calculates the real-time additional speed instructions of the trolley and the rope within 10 milliseconds based on the received data and the preset anti-sway algorithm, and sends it to the mechanism control module through the CAN bus; S4: After receiving the command, the mechanism control module immediately adjusts the motor output to achieve stable control of the grab bucket; S5: The main controller module continuously monitors the working status of each module and automatically takes corresponding measures when an abnormality is detected, such as switching to the backup CPU, issuing an alarm, etc.

Images