Multi-sensor cooperative control method and system for floating support installation

Abstract

The invention provides a multi-sensor cooperative control method and system for floating support installation. The method comprises the following steps: S1, collecting real-time data from a sensor array; s2, dynamic environment features are extracted from the unified feature sequence, a real-time marine environment model is constructed, and disturbance parameters at the next moment are predicted based on the real-time marine environment model; s3, constructing a real-time mechanical model of the floating support; s4, transmitting the optimal control signal to a floating support execution system, and driving the floating support to carry out dynamic adjustment; and S5, updating the optimal control signal for optimization. According to the method, a complete closed loop from environment perception and attitude optimization to dynamic control in the floating support installation process is achieved, the accuracy and stability of floating support installation are remarkably improved, and meanwhile the method has the high adaptability to the dynamic marine environment.

Description

Technical Field

[0001] The present invention belongs to the field of multi-sensor collaborative control, and in particular relates to a multi-sensor collaborative control method and system for floating installation. Background Art

[0002] In marine engineering, floating installation is a complex and critical technology, which is widely used in the transportation, positioning and fixing of floating wind power foundations, deep-sea oil platform modules and large underwater structures. This process needs to overcome many challenges in the marine environment, such as the dynamic changes of water flow, waves, wind force, and the complexity of the seabed topography. These factors not only put forward extremely high requirements on the stability and precise positioning of floating, but also have a profound impact on the efficiency, safety and cost-effectiveness of the entire installation process.

[0003] Current float installation technologies rely primarily on manual operation and single sensor control strategies. For example, traditional systems typically use water flow sensors or inertial measurement units (IMUs) to monitor the dynamic state of the float during installation. However, these methods have significant limitations: Functional limitations of a single sensor: A single sensor cannot fully reflect the environmental complexity of the floating installation process. For example, a water flow sensor can monitor the flow rate, but it is difficult to provide attitude information; an IMU can monitor the attitude, but its perception of environmental interference is insufficient.

[0004] Poor real-time adaptability to environmental changes: The marine environment is dynamic and highly unpredictable, and existing technologies cannot effectively model and predict the changing trends of water currents and waves. This lack of environmental perception and prediction capabilities often leads to increased installation deviations and mission failures.

[0005] Lack of systematic collaborative control: Existing technologies are mostly single-point optimization, which makes it difficult to effectively integrate data from multiple sensors, resulting in insufficient accuracy in floating attitude adjustment and susceptibility to sudden environmental disturbances.

[0006] In the face of the above problems, some scholars have tried to apply multi-sensor fusion technology and artificial intelligence to floating installation scenarios in recent years. However, these attempts often rely too much on a single technical point (such as data fusion algorithm), fail to establish a complete closed-loop solution combining dynamic environment prediction, mechanical modeling and control optimization, and fail to fully solve the accuracy and stability problems in the floating installation process. Summary of the invention

[0007] The purpose of the present invention is to propose a multi-sensor collaborative control method and system for floating installation, which comprehensively solves the problems of precise control and stability during floating installation by real-time marine environment prediction and mechanical state optimization for multi-sensor collaborative control.

[0008] In order to achieve the above object, a multi-sensor cooperative control method for floating installation is provided in a first aspect of the present invention, the method comprising: S1, collecting real-time data from the sensor array, preprocessing the real-time data and establishing a multimodal mapping matrix, mapping the preprocessed data to a unified feature space through the multimodal mapping matrix to obtain a unified feature sequence; S2. Extract dynamic environmental features from the unified feature sequence, construct a real-time ocean environment model that combines time series trends and spatial correlations, and predict disturbance parameters at the next moment based on the real-time ocean environment model; wherein the disturbance parameters include flow velocity, wave intensity and direction at the next time step; S3. Based on the real-time ocean environment model and the predicted disturbance parameters, combined with the unified feature sequence, a real-time mechanical model of the float is constructed, and the real-time force state and attitude parameters of the float are estimated by combining the disturbance parameters in the dynamic ocean environment; wherein the force state is the total force of the float, that is, the total force of the float in the three directions of x, y, and z; the attitude parameter is the attitude parameter of the float, that is, the rotation angle of the float in the three directions of x, y, and z; S4, designing an optimization objective function according to the estimated real-time force state and attitude parameters of the float and the predicted disturbance parameters at the next moment, calculating the optimal control signal of the float, and transmitting the optimal control signal to the float execution system to drive the float to perform dynamic adjustment; S5. Monitor and collect the actual force and posture after the execution of the optimal control signal in real time, and compare the actual force and posture after the execution with the target force and posture, dynamically generate a compensation signal, and correct the compensation signal based on the comparison result of the actual force and posture parameters after the execution with the target force and posture parameters, update the optimal control signal according to the corrected compensation signal for optimization, and finally obtain the final force state and posture parameters, and compare the final stable force state and posture parameters to verify whether the target state is reached. If the error exceeds the tolerance range, dynamic feedback adjustment is performed.

[0009] Further, the sensor array includes a flow velocity sensor, a water depth pressure sensor, a sonar sensor and an inertial measurement unit; The preprocessing includes performing time sequence consistency processing, specific noise filtering, standardization processing and mapping processing on the real-time data.

[0010] Furthermore, the S2 specifically includes: According to the inhomogeneity of multimodal data in the marine environment, a weighted spatial feature mapping function is designed to dynamically adjust the contributions of different modes; A real-time ocean environment model is constructed based on the weighted spatial feature mapping function, and the hidden state, feature-to-hidden state mapping matrix and state transfer matrix are output; The disturbance parameters of the next time step are predicted according to the hidden state, and the fluid continuity constraint is introduced into the real-time ocean environment model to ensure the physical rationality of the prediction.

[0011] Furthermore, the weighted spatial feature mapping function realizes dynamic adjustment of contributions of different modes by introducing a sparse regularization term; The real-time ocean environment model uses a state transfer formula to capture the dynamic change trend of the time series, while adding specific disturbance terms of the ocean environment to output a hidden state; Based on the hidden state of the output, use a recurrent neural network to predict the next time step The disturbance parameter.

[0012] Furthermore, the S3 specifically includes: Design basic mechanical models; Adding anti-torsion performance constraints and fluid continuity regularization for floating installation to the basic mechanical model; wherein the anti-torsion performance constraints are used to limit the amount of change in the rotation angle, and the fluid continuity regularization is used to constrain its divergence to meet the fluid continuity condition; Design the overall optimization goal of the mechanical model.

[0013] Furthermore, the basic mechanical model is the total force of the real-time floatation, and the total force of the real-time floatation includes buoyancy, gravity, dynamically generated control signals and drag force; wherein the dynamically generated control signals represent the floatation in Directional forces; The anti-torsion performance constraint is used to limit the change in rotation angle; The fluid continuity regularization is used to constrain its divergence to satisfy the fluid continuity condition; The total force and attitude parameters of real-time floatation are iteratively solved through numerical optimization methods, and the force state and attitude parameters are dynamically updated.

[0014] Furthermore, the S4 specifically includes: Design control optimization objective function; Use numerical optimization methods to solve the optimal control signal; Output optimized control signals, adjust the force state and posture of the float in real time, transmit the control signals to the float execution system, and drive the float to make dynamic adjustments; The control optimization objective function is used to achieve force balance and attitude adjustment. The force balance means that the total force is closest to the ideal force state, and the attitude adjustment means that the attitude parameters of the real-time floating are closest to the target attitude.

[0015] Furthermore, a dynamic disturbance compensation term and an energy consumption limitation term are added to the design control optimization objective function; the dynamic disturbance compensation term is used to reduce the influence of floatation response delay, and the energy consumption limitation term is used to limit the intensity of the control signal.

[0016] Further, in S5, the optimal control signal is executed, and the thruster and the attitude adjustment device act on the floatation, and the thruster adjusts the floatation according to the optimal control signal. The attitude adjustment device adjusts the rotation angle of the float according to the target attitude; the compensation signal is compensated by superimposing on the original control signal; By correcting the compensation signal and generating a correction vector of environmental disturbance based on the prediction of the ocean environment model, the dynamic disturbance correction term is designed to reduce the impact of sudden environmental disturbances on the floating state.

[0017] A multi-sensor collaborative control system for floating installation, the system comprising: In another aspect of the present invention, a multi-sensor cooperative control system for floating installation is provided, the system comprising: A real-time data acquisition unit is used to collect real-time data from the sensor array, pre-process the real-time data and establish a multi-modal mapping matrix, and map the pre-processed data to a unified feature space through the multi-modal mapping matrix to obtain a unified feature sequence; An environmental prediction unit is used to extract dynamic environmental features from a unified feature sequence, construct a real-time ocean environment model that combines time series trends with spatial correlations, and predict disturbance parameters at the next moment based on the real-time ocean environment model; wherein the disturbance parameters include flow velocity, wave intensity and direction at the next time step; The float installation analysis unit is used to construct a real-time mechanical model of the float based on the real-time marine environment model and the predicted disturbance parameters in combination with a unified feature sequence, and estimate the real-time force state and attitude parameters of the float by combining the disturbance parameters in the dynamic marine environment; wherein the force state is the total force of the float, that is, the total force of the float in the three directions of x, y, and z; the attitude parameter is the attitude parameter of the float, that is, the rotation angle of the float in the three directions of x, y, and z; An adjustment control unit is installed to design an optimization objective function based on the estimated real-time force state and attitude parameters of the float and the predicted disturbance parameters at the next moment, calculate the optimal control signal of the float, and transmit the optimal control signal to the float execution system to drive the float to perform dynamic adjustment; A control optimization unit is installed to monitor and collect the actual force and attitude after the execution of the optimal control signal in real time, and compare the actual force and attitude after the execution with the target force and attitude, dynamically generate a compensation signal, and correct the compensation signal based on the comparison result of the actual force and attitude parameters after the execution with the target force and attitude parameters, update the optimal control signal according to the corrected compensation signal for optimization, and finally obtain the final force state and attitude parameters, and compare the final stable force state and attitude parameters to verify whether the target state is reached. If the error exceeds the tolerance range, dynamic feedback adjustment is performed.

[0018] The beneficial technical effects of the present invention are at least as follows: (1) The present invention constructs a dynamic ocean environment model through data fusion of multiple sensors (such as water flow sensors, sonar sensors, pressure sensors, etc.), and predicts the changing trends of interference factors such as water flow and waves in real time. This technology solves the problem of poor adaptability to dynamic environmental changes in the prior art, ensuring that the environmental information during the floating installation process is comprehensive and accurate.

[0019] (2) The present invention combines real-time environmental perception data to establish an accurate mechanical model of the float, dynamically estimates its force state (such as buoyancy, gravity, and drag), and adjusts the float attitude in real time through model predictive control (MPC). This technology overcomes the problem of lack of systematic mechanical optimization in traditional methods and achieves precise control of the float attitude.

[0020] (3) A multi-sensor collaborative control mechanism based on the extended Kalman filter (EKF) and adaptive control algorithm was designed to achieve efficient fusion and feedback of different sensor data. This mechanism improves the robustness of the system and solves the shortcomings of low sensor data utilization and insufficient anti-interference ability in the existing technology. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] The present invention is further described using the accompanying drawings, but the embodiments in the accompanying drawings do not constitute any limitation to the present invention. A person skilled in the art can obtain other drawings based on the following drawings without creative work.

[0022] Figure 1 This is a flow chart of a multi-sensor collaborative control method for floating installation according to an embodiment of the present invention.

[0023] Figure 2 This is a framework diagram of a multi-sensor collaborative control system for floating installation according to an embodiment of the present invention. DETAILED DESCRIPTION

[0024] Embodiments of the present invention are described in detail below, examples of which are shown in the accompanying drawings, wherein the same or similar reference numerals throughout represent the same or similar elements or elements having the same or similar functions. The embodiments described below with reference to the accompanying drawings are exemplary and are only used to explain the present invention, and cannot be understood as limiting the present invention.

[0025] like Figure 1 As shown, an embodiment of the present invention provides a multi-sensor collaborative control method for floating installation, the method comprising: S1. Collect real-time data from the sensor array, pre-process the real-time data and establish a multimodal mapping matrix, map the pre-processed data to a unified feature space through the multimodal mapping matrix, and obtain a unified feature sequence.

[0026] Specifically, real-time data is collected from the sensor array and preprocessed through alignment, denoising, modal mapping and other methods to provide high-quality input data for subsequent environmental modeling. The floating installation process involves a variety of heterogeneous sensors, whose outputs have inconsistent time steps, formats and noise characteristics. Therefore, the goal of this step is to generate consistent and reliable time series data through innovative specific modal processing and feature space construction methods.

[0027] The input is the real-time output data of the sensor array. ,in is the time step, Indicates the type of sensor (such as flow sensor, water depth pressure sensor, sonar sensor, and inertial measurement unit).

[0028] Will Mapping to a time series matrix , is the number of sensors, The data dimension of each sensor output (such as a three-dimensional vector of flow velocity). For example, the output of a flow velocity sensor is , the sonar is a three-dimensional point cloud data matrix.

[0029] Furthermore, a sensor synchronization module is defined to align asynchronous data to a unified time step by interpolation. ,make sure The timestamps are exactly the same. Interpolation formula: ; in, and are the sampling values ​​at the last two time points, To align the time.

[0030] Furthermore, spatial directional filtering is used on the sonar data to divide the signal into two parts: background noise and reflection data: For the point cloud matrix , by defining the direction vector and reflection points Angle : ; in, is the sonar sensor position, is the detection direction of the sensor, only Points less than the set threshold .

[0031] For the velocity data, Fourier transform is used to remove the interference signal of a specific frequency and the conversion formula is defined: ; Then only keep the low-frequency signal for inverse transformation .

[0032] Furthermore, in order to unify the data format and scale, a multimodal mapping matrix is ​​established , all sensor outputs Mapping to a unified feature space: ; Defined by feature selection matrix, such as selecting intensity for velocity , extract the point cloud density of the sonar, and finally get a unified feature sequence ,in To unify the feature space dimension.

[0033] right Standardization is performed to ensure that the features have a distribution with mean 0 and variance 1.

[0034] Finally, the output is a unified time series after alignment, denoising, mapping and normalization. , as input for subsequent steps, with high quality and consistency.

[0035] S2. Extract dynamic environmental features from the unified feature sequence, construct a real-time ocean environment model that combines time series trends and spatial correlations, and predict the disturbance parameters at the next moment based on the real-time ocean environment model; wherein the disturbance parameters include flow velocity, wave intensity and direction at the next time step.

[0036] Specifically, initialize the hidden state , is the dimension of the hidden state, which is used to capture dynamic change trends. The initial value of the hidden state can be set to a zero vector, that is, .

[0037] Furthermore, an improved weighted spatial feature mapping method is proposed to address the inhomogeneity of multimodal data in the marine environment. , by introducing innovative weight regularization terms, dynamic adjustment of contributions of different modes is achieved: ; in, is the weight matrix of the feature map, is the output feature dimension; is the bias vector; is the sparse regularization term, Sparse coefficient, used to improve the contribution of low-dimensional significant features to the final prediction; is a nonlinear activation function (such as ReLU or Tanh). Output feature map Used for subsequent dynamic transfer modeling.

[0038] Furthermore, the state transfer formula is used to capture the dynamic trend of the time series, while adding specific disturbance terms of the marine environment , the modeling formula is as follows: ; in, is the mapping matrix from features to hidden states; is the state transfer matrix; is the state bias vector; is the innovative disturbance term, is the perturbation weight, Constructed from physical features extracted from multimodal data (e.g., gradient field of flow velocity or rate of change of wave frequency). Output hidden state , as the dynamic feature of the next time step.

[0039] Furthermore, based on the hidden state , predict the next time step The disturbance parameter , and introduce fluid continuity constraints into the output prediction model to ensure the physical rationality of the prediction: ; in, is the output weight matrix; is the output bias; is the fluid continuity constraint, Represents the divergence of the velocity field and controls the prediction results to meet the physical continuity conditions.

[0040] Finally, output the environment dynamic model , used to describe the dynamic changes in the ocean environment. Output predicted disturbance parameters , including the flow velocity, wave intensity and direction of the next time step, which is used for subsequent mechanical state estimation.

[0041] Through innovative dynamic time series modeling and regularized design targeting the characteristics of the marine environment, this solution not only captures the regularity of the time series, but also introduces additional constraints with clear physical meanings (such as fluid continuity and disturbance terms), making the prediction results more suitable for the complex environment of floating installation scenarios.

[0042] S3. Based on the real-time ocean environment model and predicted disturbance parameters, combined with a unified feature sequence, a real-time mechanical model of the float is constructed. The real-time force state and attitude parameters of the float are estimated by combining the disturbance parameters in the dynamic ocean environment.

[0043] Specifically, initialize the variables: : Total force of floatation in three directions; : The attitude parameter of the float, describing the rotation angles in three directions.

[0044] Furthermore, the design of the basic mechanical model: Total force of floatation It consists of the following forces: ; buoyancy : Based on floatation submerged volume , water depth data from the sensor calculate; gravity : By floatation mass and gravitational acceleration are determined; Drag : With the predicted flow rate Proportional to the flow area and drag coefficient Sure; Control : Subsequent steps are dynamically generated to offset environmental disturbances.

[0045] Furthermore, two innovative constraints for floating installation scenarios are added to the mechanical model: Torsional performance constraints: The change of the floating attitude must meet the anti-torsion design and limit the change of the rotation angle : ; in , is the angular velocity, is the weight coefficient.

[0046] Fluid Continuity Regularization: Using the velocity field predicted in step S2 , constraining its divergence to satisfy the fluid continuity condition: ; This regularization ensures that the mechanical estimation results are consistent with the actual fluid behavior.

[0047] Furthermore, in order to ensure the stability and force balance of the float, the overall optimization objectives of the design are: ; in, It is an ideal force balance state; and is the constraint weight coefficient. It is solved iteratively by numerical optimization methods (such as gradient descent method) and , dynamically update the force state and posture parameters.

[0048] Finally, output the estimated total floatation force , is the real-time force state of the float in a dynamic environment; output attitude parameters , describes the rotation angle of the float, and provides the core input for the dynamic control optimization in step S4. This result effectively adapts to the dynamic disturbance of the complex environment and provides high-precision mechanical data support for subsequent optimization control.

[0049] In this step, a mechanical estimation model suitable for floating installation scenarios is constructed by innovatively combining fluid continuity regularization and anti-torsion performance constraints.

[0050] S4. Design an optimization objective function based on the estimated real-time force state and attitude parameters of the float and the predicted disturbance parameters at the next moment, calculate the optimal control signal of the float, and transmit the optimal control signal to the float execution system to drive the float to perform dynamic adjustment.

[0051] Specifically, : Dynamically generated control signal, indicating that the float is Directional forces; : Target attitude, defines the ideal rotation angle for floating.

[0052] Furthermore, in order to meet the requirements of floatation force balance and attitude control, the control optimization objective function is designed to give priority to the following two points: Force balance: total force Should be close to the ideal stress state ; Posture adjustment: real-time posture Should tend to the target posture .

[0053] Furthermore, based on the above requirements, the core part of the optimization objective function is: ; in, , which has been defined in step 3, represents the ideal balanced force state; It is the weight coefficient of attitude adjustment, which is used to adjust the priority of force balance and attitude control.

[0054] Furthermore, in order to meet the special requirements of floating installation, the following innovative regularization terms are added to the optimization objective function: Dynamic disturbance compensation term: dynamic environmental disturbance predicted in step 2 , design compensation mechanism to reduce the impact of floatation response delay: ; in is the current sensor data, is the compensation strength coefficient.

[0055] Energy consumption limit item: To reduce the energy consumption of the control system, limit the strength of the control signal: ; in is the energy consumption limit coefficient, used to constrain size.

[0056] The final optimization objective function is: ; Control signal generation method Use numerical optimization methods (such as gradient descent) to solve the optimal control signal : Initial Value: ; Update rules: ; in is the learning rate, is the gradient of the objective function, and each part of the gradient is decomposed and calculated according to the above formula.

[0057] Finally, the optimized control signal is output , used to adjust the force state and attitude of the float in real time; transmit the control signal to the float execution system (such as the propulsion device, attitude adjustment mechanism) to drive the float to make dynamic adjustments; and further optimize the dynamic stability of the float through subsequent feedback closed-loop control (step S5).

[0058] In this step, an innovative control optimization objective function including dynamic disturbance compensation and energy consumption limitation is designed to generate the optimal control signal that meets the requirements of floating installation.

[0059] S5. Monitor and collect the actual force and posture after the execution of the optimal control signal in real time, and compare the actual force and posture after the execution with the target force and posture, dynamically generate a compensation signal, and correct the compensation signal based on the comparison result of the actual force and posture parameters after the execution with the target force and posture parameters, update the optimal control signal according to the corrected compensation signal for optimization, and finally obtain the final force state and posture parameters, and compare the final stable force state and posture parameters to verify whether the target state is reached. If the error exceeds the tolerance range, dynamic feedback adjustment is performed.

[0060] Specifically, the control execution and feedback closed loop is the core execution part of the entire patent solution, and the control signal generated by applying step S4 Adjust the movement state of the float and monitor the actual force in real time and posture , compare it with the target state and dynamically generate a compensation signal The scheme designs an innovative feedback regulation strategy, which improves the adaptability and control accuracy of the system by introducing the correction of environmental disturbance terms and dynamic adjustment of feedback gain, and finally forms an efficient closed-loop control mechanism.

[0061] Variable definition: : The actual stress state of the float is measured in real time by the sensor; : The actual attitude of the float is monitored in real time by the inertial measurement unit (IMU); : force error vector; : attitude error vector.

[0062] Furthermore, the execution control signal , acting on the float through the thrusters and attitude adjustment device: Propeller according to Adjust floatation Direction of movement; The attitude adjustment device adjusts the target attitude according to the target attitude. Adjust the rotation angle of the float.

[0063] Real-time monitoring of the actual force of floatation and posture , update the current status.

[0064] Furthermore, according to the real-time status, the current force error is calculated and attitude error , and generate compensation signal : ; in, is the force feedback gain matrix, is the attitude feedback gain matrix; Compensation is achieved by superimposing on the original control signal.

[0065] Furthermore, a dynamic disturbance correction term is introduced to correct the feedback error based on the environmental model in step 2 to reduce the impact of dynamic changes in the ocean environment on control accuracy: ; in, is the correction vector of the environmental disturbance, which is generated by the dynamic environment model prediction in step 2; is the correction weight coefficient, which is used to balance the strength of error compensation and disturbance correction. Through the correction term, the impact of sudden environmental disturbances on the floating state can be further reduced.

[0066] Furthermore, the updated control signal is: ; implement And loop feedback until the error meets the tolerance range: ; in and are tolerance thresholds, representing the acceptable range of force error and posture error respectively.

[0067] Finally, the final stable stress state is output and attitude parameters , verify whether the floatation has reached the target state; if the error exceeds the tolerance range, the system continues to dynamically feedback and adjust to form a closed-loop optimization to ensure stable installation of the floatation.

[0068] This step innovatively combines dynamic disturbance correction with feedback error compensation to form an adaptive closed-loop control mechanism for floating installation scenarios. While reducing unnecessary formulas, correction items for environmental disturbances are designed, providing the patented solution with powerful dynamic response capabilities and system stability guarantees.

[0069] like Figure 2 As shown, in another embodiment of the present invention, a multi-sensor cooperative control system for floating installation is provided, the system comprising: The real-time data acquisition unit 1011 is used to collect real-time data from the sensor array, pre-process the real-time data and establish a multi-modal mapping matrix, and map the pre-processed data to a unified feature space through the multi-modal mapping matrix to obtain a unified feature sequence; The environmental prediction unit 1012 is used to extract dynamic environmental features from the unified feature sequence, construct a real-time ocean environment model that combines time series trends and spatial correlations, and predict disturbance parameters at the next moment based on the real-time ocean environment model; wherein the disturbance parameters include flow velocity, wave intensity and direction at the next time step; The floating installation analysis unit 1013 is used to construct a real-time mechanical model of the floating device based on the real-time ocean environment model and the predicted disturbance parameters in combination with a unified feature sequence, and estimate the real-time force state and attitude parameters of the floating device by combining the disturbance parameters in the dynamic ocean environment; wherein the force state is the total force of the floating device, that is, the total force of the floating device in the three directions of x, y, and z; the attitude parameter is the attitude parameter of the floating device, that is, the rotation angle of the floating device in the three directions of x, y, and z; An adjustment control unit 1014 is installed, which is used to design an optimization objective function according to the estimated real-time force state and attitude parameters of the float and the predicted disturbance parameters at the next moment, calculate the optimal control signal of the float, and transmit the optimal control signal to the float execution system to drive the float to perform dynamic adjustment; A control optimization unit 1015 is installed to monitor and collect the actual force and attitude after the execution of the optimal control signal in real time, and compare the actual force and attitude after the execution with the target force and attitude, dynamically generate a compensation signal, and correct the compensation signal based on the comparison result of the actual force and attitude parameters after the execution with the target force and attitude parameters, update the optimal control signal according to the corrected compensation signal for optimization, and finally obtain the final force state and attitude parameters, and compare the final stable force state and attitude parameters to verify whether the target state is reached. If the error exceeds the tolerance range, dynamic feedback adjustment is performed.

[0070] In summary, the present invention realizes a complete closed loop from environmental perception, posture optimization to dynamic control during the floating installation process, significantly improves the accuracy and stability of the floating installation, and at the same time has high adaptability to the dynamic marine environment.

[0071] It should be noted that the workflow described above is merely illustrative and does not limit the scope of protection of the present invention. In practical applications, technicians in this field can select part or all of them according to actual needs to achieve the purpose of the present embodiment, and no limitation is made here.

[0072] In addition, for technical details that are not described in detail in this embodiment, reference can be made to the parameter operation method provided in any embodiment of the present invention, and will not be repeated here.

[0073] It should be noted that, in this article, the terms "include", "comprises" or any other variations thereof are intended to cover non-exclusive inclusion, so that a process, method, article or system including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or system. In the absence of further restrictions, an element defined by the sentence "comprises a ..." does not exclude the existence of other identical elements in the process, method, article or system including the element.

[0074] The serial numbers of the above embodiments of the present invention are only for description and do not represent the advantages or disadvantages of the embodiments.

[0075] Through the description of the above implementation methods, those skilled in the art can clearly understand that the above-mentioned embodiment methods can be implemented by means of software plus a necessary general hardware platform, and of course by hardware, but in many cases the former is a better implementation method. Based on such an understanding, the technical solution of the present invention, or the part that contributes to the prior art, can be embodied in the form of a software product, which is stored in a storage medium (such as a read-only memory / random access memory, a magnetic disk, or an optical disk), and includes a number of instructions for enabling a terminal device (which can be a mobile phone, a computer, a server, an air conditioner, or a network device, etc.) to execute the methods described in each embodiment of the present invention.

[0076] The above are only preferred embodiments of the present invention, and are not intended to limit the patent scope of the present invention. Any equivalent structure or equivalent process transformation made using the contents of the present invention specification and drawings, or directly or indirectly applied in other related technical fields, are also included in the patent protection scope of the present invention.

Claims

1. A multi-sensor collaborative control method for floating installation, characterized in that: The method comprises: S1, collecting real-time data from the sensor array, preprocessing the real-time data and establishing a multimodal mapping matrix, mapping the preprocessed data to a unified feature space through the multimodal mapping matrix to obtain a unified feature sequence; S2. Extract dynamic environmental features from the unified feature sequence, construct a real-time ocean environment model that combines time series trends and spatial correlations, and predict disturbance parameters at the next moment based on the real-time ocean environment model; wherein the disturbance parameters include flow velocity, wave intensity and direction at the next time step; S3. Based on the real-time ocean environment model and the predicted disturbance parameters, combined with the unified feature sequence, a real-time mechanical model of the float is constructed, and the real-time force state and attitude parameters of the float are estimated by combining the disturbance parameters in the dynamic ocean environment; wherein the force state is the total force of the float, that is, the total force of the float in the three directions of x, y, and z; the attitude parameter is the attitude parameter of the float, that is, the rotation angle of the float in the three directions of x, y, and z; S4, designing an optimization objective function according to the estimated real-time force state and attitude parameters of the float and the predicted disturbance parameters at the next moment, calculating the optimal control signal of the float, and transmitting the optimal control signal to the float execution system to drive the float to perform dynamic adjustment; S5. Monitor and collect the actual force and posture after the execution of the optimal control signal in real time, and compare the actual force and posture after the execution with the target force and posture, dynamically generate a compensation signal, and correct the compensation signal based on the comparison result of the actual force and posture parameters after the execution with the target force and posture parameters, update the optimal control signal according to the corrected compensation signal for optimization, and finally obtain the final force state and posture parameters, and compare the final stable force state and posture parameters to verify whether the target state is reached. If the error exceeds the tolerance range, dynamic feedback adjustment is performed.

2. A multi-sensor cooperative control method for floating installation according to claim 1, characterized in that: The sensor array includes a flow velocity sensor, a water depth pressure sensor, a sonar sensor and an inertial measurement unit; The preprocessing includes performing time sequence consistency processing, specific noise filtering, standardization processing and mapping processing on the real-time data.

3. A multi-sensor cooperative control method for floating installation according to claim 1, characterized in that: The S2 specifically includes: According to the inhomogeneity of multimodal data in the marine environment, a weighted spatial feature mapping function is designed to dynamically adjust the contributions of different modes; A real-time ocean environment model is constructed based on the weighted spatial feature mapping function, and the hidden state, feature-to-hidden state mapping matrix and state transfer matrix are output; The disturbance parameters of the next time step are predicted according to the hidden state, and the fluid continuity constraint is introduced into the real-time ocean environment model to ensure the physical rationality of the prediction.

4. A multi-sensor cooperative control method for floating installation according to claim 3, characterized in that: The weighted spatial feature mapping function realizes dynamic adjustment of contributions of different modes by introducing sparse regularization terms; The real-time ocean environment model uses a state transfer formula to capture the dynamic change trend of the time series, while adding specific disturbance terms of the ocean environment to output a hidden state; Based on the hidden state of the output, use a recurrent neural network to predict the next time step The disturbance parameter.

5. The multi-sensor cooperative control method for floating installation according to claim 1, characterized in that: The S3 specifically includes: Design basic mechanical models; Adding anti-torsion performance constraints and fluid continuity regularization for floating installation to the basic mechanical model; wherein the anti-torsion performance constraints are used to limit the amount of change in the rotation angle, and the fluid continuity regularization is used to constrain its divergence to meet the fluid continuity condition; Design the overall optimization goal of the mechanical model.

6. A multi-sensor cooperative control method for floating installation according to claim 5, characterized in that: The basic mechanical model is the total force of the real-time floatation, which includes buoyancy, gravity, dynamically generated control signals and drag force; wherein the dynamically generated control signals represent the floatation Directional forces; The anti-torsion performance constraint is used to limit the change in rotation angle; The fluid continuity regularization is used to constrain its divergence to satisfy the fluid continuity condition; The total force and attitude parameters of real-time floatation are iteratively solved through numerical optimization methods, and the force state and attitude parameters are dynamically updated.

7. A multi-sensor cooperative control method for floating installation according to claim 6, characterized in that: The S4 specifically includes: Design control optimization objective function; Use numerical optimization methods to solve the optimal control signal; Output optimized control signals, adjust the force state and posture of the float in real time, transmit the control signals to the float execution system, and drive the float to make dynamic adjustments; The control optimization objective function is used to achieve force balance and attitude adjustment. The force balance means that the total force is closest to the ideal force state, and the attitude adjustment means that the attitude parameters of the real-time floating are closest to the target attitude.

8. A multi-sensor cooperative control method for floating installation according to claim 7, characterized in that: A dynamic disturbance compensation term and an energy consumption limitation term are added to the design control optimization objective function; the dynamic disturbance compensation term is used to reduce the influence of floatation response delay, and the energy consumption limitation term is used to limit the intensity of the control signal.

9. The multi-sensor cooperative control method for floating installation according to claim 7, characterized in that: In S5, the optimal control signal is executed, and the thruster and the attitude adjustment device act on the float. The thruster adjusts the float according to the optimal control signal. The attitude adjustment device adjusts the rotation angle of the float according to the target attitude; the compensation signal is compensated by superimposing on the original control signal; By correcting the compensation signal and generating a correction vector of environmental disturbance based on the prediction of the ocean environment model, the dynamic disturbance correction term is designed to reduce the impact of sudden environmental disturbances on the floating state.

10. A multi-sensor cooperative control system for floating installation, characterized in that: The system comprises: A real-time data acquisition unit is used to collect real-time data from the sensor array, pre-process the real-time data and establish a multi-modal mapping matrix, and map the pre-processed data to a unified feature space through the multi-modal mapping matrix to obtain a unified feature sequence; An environmental prediction unit is used to extract dynamic environmental features from a unified feature sequence, construct a real-time ocean environment model that combines time series trends with spatial correlations, and predict disturbance parameters at the next moment based on the real-time ocean environment model; wherein the disturbance parameters include flow velocity, wave intensity and direction at the next time step; The float installation analysis unit is used to construct a real-time mechanical model of the float based on the real-time marine environment model and the predicted disturbance parameters in combination with a unified feature sequence, and estimate the real-time force state and attitude parameters of the float by combining the disturbance parameters in the dynamic marine environment; wherein the force state is the total force of the float, that is, the total force of the float in the three directions of x, y, and z; the attitude parameter is the attitude parameter of the float, that is, the rotation angle of the float in the three directions of x, y, and z; An adjustment control unit is installed to design an optimization objective function based on the estimated real-time force state and attitude parameters of the float and the predicted disturbance parameters at the next moment, calculate the optimal control signal of the float, and transmit the optimal control signal to the float execution system to drive the float to perform dynamic adjustment; A control optimization unit is installed to monitor and collect the actual force and attitude after the execution of the optimal control signal in real time, and compare the actual force and attitude after the execution with the target force and attitude, dynamically generate a compensation signal, and correct the compensation signal based on the comparison result of the actual force and attitude parameters after the execution with the target force and attitude parameters, update the optimal control signal according to the corrected compensation signal for optimization, and finally obtain the final force state and attitude parameters, and compare the final stable force state and attitude parameters to verify whether the target state is reached. If the error exceeds the tolerance range, dynamic feedback adjustment is performed.

Images