Offshore wind farm installation vessel dynamic positioning operation training simulation system

Abstract

The application relates to a kind of offshore wind power installation ship power positioning operation training simulation systems, including sensor and positioning module, motion model simulation module, man-machine interaction module, controller module, navigation planning module, propeller distribution module, network interface module, record and playback module and evaluation module. The initial state of the ship and the navigation operation environment can be simulated by the trainer, and the navigation and power positioning operation training of the simulated wind power installation ship can be realized by using the simulated bridge. Compared with the prior art, the application can be used for wind power installation ship power positioning operator training, examination and daily training, thereby improving the operation level of the power positioning operator and improving the efficiency of offshore operation. At the same time, it can also break the monopoly situation of foreign power positioning training.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of offshore wind power installation ship power positioning control, in particular to a kind of offshore wind power installation ship power positioning operation training simulation system. BACKGROUND

[0002] Offshore wind power equipment installation needs a special engineering ship-wind power installation ship. Offshore wind power installation ship operating environment is complex, and power positioning and crane operators must be certified, and the training of power positioning operators is currently monopolized by foreign countries, with high training prices and long training cycles, and real ships cannot provide similar operation training. Therefore, a power positioning operation training simulation system suitable for offshore wind power installation ship needs to be developed for wind power installation ship power positioning operator training, examination and daily training, to improve the operation level of power positioning operators and improve offshore operation efficiency, and also to break the monopoly of foreign power positioning training. SUMMARY

[0003] The purpose of the present application is to overcome the defects of the prior art and provide a kind of offshore wind power installation ship power positioning operation training simulation system.

[0004] The purpose of the present application can be achieved by the following technical solutions:

[0005] A kind of offshore wind power installation ship power positioning operation training simulation system, including sensor and positioning module, motion model simulation module, man-machine interaction module, controller module, navigation planning module and propeller distribution module;

[0006] The sensor and positioning module is used to simulate the position of wind power installation ship and the data of nearby sea conditions;

[0007] The motion model simulation module is used to construct a ship motion model by combining the fixed parameters of the simulated wind power installation ship;The fixed parameters include ship length, ship width, draft, superstructure wind area and propeller distribution;

[0008] The man-machine interaction module includes a simulated bridge display unit, and the motion model simulation module transmits data to the simulated bridge display unit of the man-machine interaction module to provide visual reference;

[0009] The simulated bridge display unit is used for operators to simulate the operation of wind power installation ship, and the operation instructions implemented by the simulation ship are received by the controller module;

[0010] The navigation planning module is used to actively read the corresponding data of the sensor and positioning module, combine the setting of the destination position and heading of the human-computer interaction module, plan the optimal path and select the appropriate sub-navigation point in real time, and send the sub-navigation point coordinate set to the controller module; the controller module reads the operation instruction and sub-navigation point coordinate set, calculates the direction and size of the thrust required by the simulated thruster of the ship through the motion control algorithm; the controller module sends the calculation result to the thruster distribution module to realize the precise control of the thruster, and then the thruster module feeds back the propulsion effect to the motion model simulation module for the correction of the ship motion mathematical model.

[0011] Further, the sensor and positioning module simulates sea state data, including manual sea state simulation mode and automatic sea state simulation mode, the manual sea state simulation mode is manually inputting wind wave parameters by the operator; the automatic sea state simulation mode is pre-downloading wind wave parameters of a specific sea area and calling in the simulation system. The two sea state simulation modes can meet the routine training and targeted training tasks of the dynamic positioning operator.

[0012] Further, the ship motion model in the motion model simulation module adopts extended Kalman filter technology, the extended Kalman filter takes wind, flow, ship position and thruster feedback as input data, compares the predicted data of the simulated ship coordinates, heading and ship motion model to obtain the difference between the two to update the ship motion model. The extended Kalman filter technology updates the model by calculating the difference between the simulated measurement data and the predicted data of the ship motion model, which is beneficial to improve the stability of the ship motion model.

[0013] Further, the navigation planning module takes the simulated ship position and heading as the initial state, and takes the actively inputted navigation end point coordinates and expected ship heading as the target state, and plans the navigation path on the positioning map view. By using image recognition algorithm, more sub-navigation points are set for the curved path on the navigation path, and fewer sub-navigation points are set for the relatively straight path. The sub-navigation point coordinate set is sent to the controller as the output data of the navigation planning module.

[0014] Further, the motion control algorithm in the controller module needs to calculate the target thrust and target direction of each thruster respectively, including the following steps:

[0015] Taking the upper limit of the thrust of the thruster and the limit of the propulsion direction as the constraint condition;

[0016] According to the forward direction, the thrust weight of the two forward thrusters is reduced, and the moment weight is increased to stabilize the heading;

[0017] The simulated annealing algorithm is executed to calculate and save multiple current optimal schemes, the current optimal schemes are taken as initial values of crossover and variation to expand the population, and the simulated annealing algorithm is continuously executed after cooling until the optimal scheme is screened out.

[0018] The motion control algorithm combines the advantages of the simulated annealing algorithm and the genetic algorithm, and has the characteristics of quickly finding a global optimal thrust distribution scheme.

[0019] Further, a deep learning BP neural network is used in the approach and retreat operation process of the dynamic positioning operation, including the following steps:

[0020] A BP neural network model is established on a QT platform;

[0021] Each changed thrust operation instruction and the corresponding system state response received by the simulation system are taken as a pair of training samples;

[0022] The optimal connection weight value is calculated by repeatedly training the connection weight value between the optimized operation instruction and the simulation system state response, and is applied to the simulation of the ship. The trained deep learning BP neural network can evolve into a neural network with prediction characteristics.

[0023] Further, the simulation system further comprises a network interface module;

[0024] The motion model simulation module and the controller module are connected with the network interface module, and are used to upload the operator's operation instruction and the motion data of the simulated ship in real time.

[0025] Further, the simulation system further comprises a recording and playback module, which is used to collect key data of this operation training, and generates a record file at the end of the operation; the human-computer interaction module can select and call the record file to reproduce the detailed operation process. After selecting to start recording, the simulation system triggers navigation data monitoring, and saves the data as a video file. After selecting to record and play back, the operator can select a specified file to realize playback on the main interface.

[0026] Further, the recording and playback module comprises a high-speed storage unit, and the recording link and the playback link are realized by using QT;

[0027] The recording link calls the avilib library under the QT environment, captures frame images of the display interface, and saves them in sequence;

[0028] The playback link uses QVideoWidget and QMediaPlayer to complete the design of the video player.

[0029] Further, the simulation system further comprises an evaluation module for automatically obtaining the operation instruction of the operator and the motion data of the simulation ship, comparing with the pre-prepared scoring standard and scoring after the training is finished.

[0030] Further, the evaluation module comprises a qualification determination and a scoring rule;

[0031] The qualification determination comprises whether to complete within the limited time, whether to meet the ship position error and whether to be stable at the target heading;

[0032] The scoring rule conforms to the following function:

[0033] S = w1 (T0-T) + w2E + w3∑(R0-R) 2

[0034] w1 + w2 + w3 = 1

[0035] In the formula, T0 is the evaluation limited operation time, T is the actual operation time of the operator, E represents a formula related to energy consumption in the operation process, ∑(R0-R) 2 is the variance of the track curve fitting degree, w1, w2 and w3 are corresponding weight coefficients respectively.

[0036] Compared with the prior art, the present application has the following beneficial effects:

[0037] 1. The present application designs a complete offshore wind power installation ship dynamic system control simulation system, which can be used for wind power installation ship dynamic positioning operator training, examination and daily training, so as to improve the operation level of the dynamic positioning operator and improve the offshore operation efficiency, and also can break the monopoly situation of foreign dynamic positioning training.

[0038] 2. The present application sets a manual sea state simulation mode and an automatic sea state simulation mode, which can meet the regular training and targeted training tasks of the dynamic positioning operator.

[0039] 3. The present application uses the extended Kalman filter technology in the ship motion model, which is beneficial to improve the stability of the ship motion model.

[0040] 4. The motion control algorithm used in the present application combines the advantages of the simulated annealing algorithm and the genetic algorithm, and has the characteristics of quickly finding the global optimal thrust distribution scheme.

[0041] 5、The simulation system can be scored objectively according to the scoring rules designed by the trainer according to the assessment requirements, so that the evaluation fluctuation caused by subjective factors can be effectively reduced, and the unification of the dynamic positioning evaluation standard can be achieved. BRIEF DESCRIPTION OF DRAWINGS

[0042] Figure 1 is a block diagram of the present application;

[0043] Figure 2 is a layout schematic diagram of the simulation bridge display unit in the embodiment of the present application;

[0044] Figure 3 is a layout schematic diagram of the simulation bridge operation panel in the embodiment of the present application;

[0045] Figure 4 is a processing flowchart of the extended Kalman filter in the embodiment of the present application;

[0046] Figure 5 is a schematic diagram of the thrust distribution strategy adopted by the simulation system in the embodiment of the present application;

[0047] Figure 6 is a deep learning flowchart of the entry point and withdrawal point dynamic positioning in the embodiment of the present application;

[0048] Figure 7 is a flow of automatic generation and review playback of operation records in the embodiment of the present application;

[0049] Figure 8 is a general flowchart of the evaluation link in the embodiment of the present application. DETAILED DESCRIPTION

[0050] The present application will be described in detail below in combination with the drawings and specific embodiments. The present embodiment is implemented on the premise of the technical solution of the present application, and detailed implementation modes and specific operation processes are given, but the protection scope of the present application is not limited to the following embodiments.

[0051] To solve the training and certification problem of the dynamic positioning operator of the offshore wind power installation ship at the present stage, the present application designs a complete offshore wind power installation ship dynamic system control simulation system for the training, assessment and daily training of the dynamic positioning operator of the offshore wind power installation ship.

[0052] As shown in Figure 1 , the offshore wind power installation ship dynamic positioning operation training simulation system of the present application includes a sensor and positioning module, a motion model simulation module, a man-machine interaction module, a controller module, a navigation planning module, a thruster distribution module, a network interface module, a record and playback module and an evaluation module.

[0053] The sensor and positioning module is used for simulating the ship position and nearby sea state data, and the motion model simulation module combines the fixed parameters of the ship (ship length, ship width, draft, superstructure wind area, and thruster distribution) to build a ship motion model.

[0054] The motion model simulation module transmits data to the display unit of the human-computer interaction module to provide visual reference. The operator's operation instructions for simulating the ship are received by the controller module through the simulation operation panel of the human-computer interaction module.

[0055] When the automatic dynamic positioning mode is selected, the navigation planning module actively reads the corresponding data of the sensor and positioning module, combines the setting of the destination position and heading of the human-computer interaction module, plans the optimal path in real time and selects appropriate navigation points, and sends the navigation point set to the controller module. The controller module reads the operation instructions and navigation point coordinate set, calculates the direction and size of the thrust that the simulated thruster of the ship needs to provide through a specific motion control algorithm. The controller module sends the calculation results to the thruster distribution module to realize precise control of the thruster, and then the thruster module feeds back the propulsion effect to the motion model simulation module for correction of the ship motion mathematical model. The motion model simulation module and the controller module are connected with the network interface module, and can upload the operator's operation instructions and the motion state of the simulated ship in real time. The uploaded data can be reproduced on the operation platform and used for training review and evaluation. After starting the recording mode, the recording and playback module collects the key data of the operation training, and generates a record file at the end of the operation. The human-computer interaction module can select to call the record file to reproduce the detailed operation process. The evaluation module automatically obtains the operator's operation instructions and the motion data of the simulated ship, compares them with the pre-prepared scoring standard, and scores after the training is completed.

[0056] It is worth noting that the simulation system in this embodiment is mainly developed in the QT environment, and in this simulation system, the wind wave data of a specific sea area needs to be saved as an.xlsx type file, and the QtXlsx tool in the QT environment is used to read the wind wave parameters.

[0057] As shown in Figure 2 , the simulation bridge display unit is developed by the QT interface library, the display interface is consistent with the interface of the actual ship dynamic positioning system, and is divided into a title bar, a work area and a side bar. The title bar presents user information, alarm prompts and other data; the main work area displays a positioning map view, the secondary work area is used to present a simulation system state view, a simulation system energy view, a simulation system thruster view, a simulation system motion trend view and the like; and the side bar presents a main menu, a motion mode selection and a display unit brightness adjustment and the like.

[0058] As shown in Figure 3As shown in the figure, the simulation console operation panel is mainly composed of simulation buttons, simulation trackballs and simulation joysticks. The operation panel of the simulation system is designed with reference to the dynamic positioning system of the actual ship, which restores the real dynamic positioning operation environment. The trackball on the simulation control panel is not used, and is replaced by the mouse of the device where the simulation system is located. The simulation operation panel is developed using the QT interface library. The simulation buttons mainly use single-click controls and double-click controls. The simulation rudder and the simulation three-axis joystick use long-click controls and single-click controls. The rotation of the two can be realized by long-clicking the left and right mouse keys.

[0059] As shown in the figure, Figure 4 The extended Kalman filter takes wind, current, ship position and thruster feedback as input data, compares the predicted data generated by the coordinate, heading and ship motion model of the simulation ship, and obtains the difference between the two to update the ship motion model. The extended Kalman filter is used to correct the ship motion model, and the stability of the model will gradually increase with the operation of the thruster.

[0060] As shown in the figure, Figure 5 The motion control algorithm model needs to calculate the target thrust and direction of each thruster respectively. Taking a four-omni-directional thruster ship as an example, there are a total of 8 variables, and the thrust upper limit and thruster direction limit are used as constraint conditions. According to the forward direction, the thrust weight of two forward thrusters is appropriately reduced, and the moment weight is increased to stabilize the heading. The simulated annealing algorithm is first executed to calculate and save multiple current optimal schemes, and then the current optimal scheme is used as the initial value of crossover and mutation to expand the population. After cooling, the simulated annealing algorithm is continued to be executed until the optimal scheme is selected. The application process of the motion control algorithm is as follows: first, use Matlab software to compile the thrust allocation strategy and save it as a.m type file, then use the Matlab cader toolbox to convert the.m file into.h and.cpp type files in C++ language. Then, through dynamic link library technology, the.cpp file is converted into a binary.dll file. Finally, the corresponding.h file and.dll file are imported into the QT interface, the function library is dynamically loaded, the preset API is called, and the thrust allocation result data is obtained.

[0061] As shown in the figure, Figure 6 As shown in the figure,

[0062] As shown in the figure, Figure 7As shown, the recording and playback module of the simulation system mainly relies on a high-speed storage unit. The recording and playback links are both implemented using QT. The recording link needs to call the avilib library in the QT environment to capture the frame images of the display interface, save them in order, and convert them into.avi format videos using an algorithm. The video file name is named "user + creation time". The playback link mainly uses QVideoWidget and QMediaPlayer to design a video player. The recording and playback module of the simulation system is designed with related controls to enhance the human-computer interaction experience. The recording and playback console contains the following main controls: record, pause, continue, stop, save, delete, playback, time tag, slider, previous button, next button, and refresh. The operation mode of the controls is similar to that of ordinary video software, which conforms to the operation habits of most users.

[0063] As shown in the figure, the evaluation module can be customized with scoring criteria, which includes three eligibility determinations and a set of scoring rules. The eligibility determinations include whether the operation is completed within the time limit, whether the position error is met, and whether the target heading is stable. If any of the determinations is no, the evaluation result is output as "unqualified". If all three determination results are yes, the evaluation result is qualified and triggers the scoring rules. The scoring rules conform to the function: Figure 8

[0064] S = w1(T0-T) + w2E + w3∑(R0-R) 2

[0065] w1 + w2 + w3 = 1

[0066] T0 is the evaluation limit operation time, T is the actual operation time of the operator, E represents a formula related to energy consumption during the operation, and ∑(R0-R) 2 is the variance of the track curve fitting degree, w1, w2, and w3 are the corresponding weight coefficients. The scoring rules are related to the operation time, energy consumption, and track curve of the operator. The less the operation time, the lower the energy consumption, and the smoother the track curve, the higher the evaluation score.

[0067] In this embodiment, the evaluation of the simulation dynamic positioning operation is set to 100 points, and the function calculation result of the scoring rules is output as the score. The simulation system uses QT to design related programs to realize evaluation scoring. The dynamic positioning training triggers a signal slot at the end, exports the actual operation time and energy consumption data during the operation, combines the limit operation time of the scoring rules, and calculates the evaluation scores of these two parts through a specific algorithm. At the same time, the QPixmap class is used to design a program to intercept the track view, separate the track curve through an image recognition algorithm, and analyze the smoothness to obtain the evaluation score of the track curve. Finally, the three evaluation scores are weighted and added to obtain the final evaluation score, which is displayed on the simulation bridge display interface. ​

[0068] The preferred embodiments of the application have been described above in detail. It should be understood that modifications and variations to the preferred embodiments could be made by those skilled in the art in light of the teachings above. It is therefore contemplated that the application can encompass other variations and modifications that fall within the scope of the claims.

Claims

1. A dynamic positioning operation training simulation system for an offshore wind turbine installation vessel, characterized in that, It includes a sensor and positioning module, a motion model simulation module, a human-computer interaction module, a controller module, a navigation planning module, and a thruster distribution module; The sensors and positioning module are used to simulate the position of the wind turbine installation vessel and nearby sea conditions. The motion model simulation module is used to construct a ship motion model by combining the fixed parameters of the simulated wind turbine installation vessel; the fixed parameters include ship length, ship width, draft, superstructure windward area, and propeller distribution; The human-computer interaction module includes a simulated driver's cab display unit. The motion model simulation module transmits data to the simulated driver's cab display unit of the human-computer interaction module to provide a visual reference. The simulated bridge display unit is used for operators to simulate operations on the wind turbine installation vessel, and the control module receives the operation instructions implemented on the simulated vessel. The navigation planning module actively reads relevant data from the sensors and positioning module, and, in conjunction with the human-machine interaction module's settings for the target ship position and heading, plans the optimal path in real time and selects suitable waypoints, sending the waypoint coordinate set to the controller module. The controller module reads the operation commands and the waypoint coordinate set, and, through a motion control algorithm, calculates the thrust direction and magnitude required by the simulated propellers of the ship. The controller module sends the calculation results to the propeller allocation module to achieve precise control of the propellers. Subsequently, the propeller module feeds back the propulsion effect to the motion model simulation module for correction of the ship's motion mathematical model. The motion control algorithm in the controller module needs to calculate the target thrust and target direction for each thruster separately, including the following steps: The upper limit of thrust and the limitation of propulsion direction of the thruster are used as constraints; Based on the forward heading, reduce the thrust weight of the two forward thrusters while increasing the torque weight to stabilize the course; The simulated annealing algorithm is executed to calculate and save multiple current better solutions. The current better solutions are used as the initial values ​​for crossover and mutation to expand the population. After cooling, the simulated annealing algorithm is executed again until the optimal solution is selected. The application of a deep learning backpropagation neural network in the entry and exit operations of dynamic positioning includes the following steps: Establish a BP neural network model on the QT platform; Each thrust-changing operation command received by the simulation system and its corresponding system state response are used as a pair of training samples. The optimal connection weights between the operational commands and the simulated system state response are calculated by repeatedly training the training samples and then applied to simulate the control of the ship. It also includes an evaluation module, which automatically acquires the operator's operating instructions and simulates the ship's motion data, compares them with pre-set scoring criteria, and scores the training after it ends. The assessment module includes pass / fail determination and scoring rules; The qualification criteria include whether the work is completed within the time limit, whether the ship's position error is met, and whether the ship is stable at the target heading. The scoring rules conform to the following function: S=w1(T0-T)+w2E+w3∑(R0-R) 2 w1 + w2 + w3 = 1 In the formula, T0 represents the assessed limited operation time, T represents the actual operation time of the operator, and E represents the formula related to energy consumption during the operation process, ∑(R0-R). 2 Let w1, w2, and w3 be the variance of the trajectory curve fitting, and w1, w2, and w3 be the corresponding weight coefficients.

2. The dynamic positioning operation training simulation system for an offshore wind turbine installation vessel according to claim 1, characterized in that, The sensor and positioning module simulates sea state data in two modes: manual and automatic. The manual sea state simulation mode involves the operator manually inputting wind and wave parameters. The automatic sea state simulation mode involves pre-downloading wind and wave parameters for a specific sea area and then calling them into the simulation system.

3. The dynamic positioning operation training simulation system for an offshore wind turbine installation vessel according to claim 1, characterized in that, The ship motion model in the motion model simulation module adopts the extended Kalman filter technology. The extended Kalman filter takes wind, current, ship position and thruster feedback as input data, compares the coordinates and heading of the simulated ship with the predicted data generated by the ship motion model, and obtains the difference between the two to update the ship motion model.

4. The dynamic positioning operation training simulation system for an offshore wind turbine installation vessel according to claim 1, characterized in that, It also includes a network interface module; Both the motion model simulation module and the controller module are connected to the network interface module to upload operator commands and simulated ship motion data in real time.

5. The dynamic positioning operation training simulation system for an offshore wind turbine installation vessel according to claim 1, characterized in that, It also includes a recording and playback module, which is used to collect key data for this operation training and generate a record file at the end of the operation; the human-computer interaction module can select to call the record file to reproduce the detailed operation process.

6. The dynamic positioning operation training simulation system for an offshore wind turbine installation vessel according to claim 5, characterized in that, The recording and playback module includes a high-speed storage unit, and both the recording and playback stages are implemented using QT. The recording process calls the avilib library in the QT environment to capture frame images of the display interface and save them in sequence. The playback process uses QVideoWidget and QMediaPlayer to design the video player.

Images