CN-121995748-A - Marine interference estimation and observer design method using multi-source sensor fusion

CN121995748ACN 121995748 ACN121995748 ACN 121995748ACN-121995748-A

Abstract

The invention is suitable for the field of offshore platform or intelligent ship control, and provides a design method of an offshore interference estimation and observer by using multi-source sensor fusion, which comprises the following steps of S1, sensor data acquisition and preprocessing, S2, interference estimation model construction, S3, state space model conversion, wherein the problem of multi-sensor time sequence deviation and coordinate system difference is solved based on a 'time alignment+coordinate conversion' structure of a sensor data acquisition and preprocessing module, the time alignment ensures that data are synchronous at a frequency of 5Hz, interference estimation delay caused by data asynchronism is avoided, and the coordinate conversion is combined with ship attitude and navigational speed to obtain accurate interference data in an absolute coordinate system, so that the accuracy of the interference data is improved by 40% -50% compared with that of the traditional preprocessing method, and reliable input is provided for a subsequent module.

Inventors

DONG YUCHEN
WANG PEIDONG
CHENG YUWEI

Assignees

陕西欧卡电子智能科技有限公司

Dates

Publication Date: 20260508
Application Date: 20251229

Claims (8)

1. The marine interference estimation and observer design method using multi-source sensor fusion is characterized by comprising the following steps: S1, sensor data acquisition and preprocessing, namely acquiring data of an Inertial Measurement Unit (IMU), a Global Positioning System (GPS), a wave sensor and a water flow sensor through a sensor data acquisition and preprocessing module, respectively performing time alignment and North East (NED) coordinate system conversion on the acquired data to obtain preprocessed data, and outputting the preprocessed data to an interference estimation model construction module; S2, constructing an interference estimation model, namely respectively constructing a wave interference model and a water flow interference model based on the received pretreatment data by an interference estimation model construction module, superposing the wave interference model and the water flow interference model to obtain a total interference initial model, and outputting the total interference initial model to a state space model module; s3, converting the state space model, namely converting the total interference initial model into a state equation describing interference dynamic change by a state space model module, and outputting the state equation and the observation equation to an interference observer design module by an observation equation describing the relation between sensor observation data and interference state; S4, estimating an interference observer, namely, based on a state equation and an observation equation, adopting a Kalman filtering algorithm by an interference observer design module, firstly predicting an interference state and a covariance matrix according to the state equation, then updating an interference estimation value by combining sensor observation data, and outputting the real-time interference estimation value to a feedforward controller design module; S5, generating a feedforward control signal, namely receiving a real-time interference estimated value by a feedforward controller design module, generating a final control signal by combining an expected control input obtained based on track planning, a target course and a speed, and outputting the final control signal to a control execution module; and S6, controlling execution, namely adjusting the rudder angle and the accelerator of the ship by the control execution module according to the final control signal, and driving the ship to execute corresponding sailing actions so as to offset external interference caused by water flow and waves.
2. The method for marine interference estimation and observer design using multi-source sensor fusion according to claim 1, wherein the "time alignment" in step S1 specifically comprises the sub-steps of: S11, determining a reference frequency, namely taking GPS output frequency (5 Hz) as a time aligned reference frequency; S12, aligning the data of the low-frequency sensor, namely aiming at a wave sensor and a water flow sensor with output frequency of 10 min/time, calculating the aligned data at any moment by adopting a linear interpolation formula, wherein the interpolation formula is that , wherein, The interpolation result at the time t is shown, For the moment of time An observation of the location; And S13, aligning high-frequency sensor data, namely performing downsampling processing according to the 5Hz frequency of the GPS for the IMU with the output frequency of 10Hz, extracting 1 IMU data point every 0.2S, and realizing time synchronization with the GPS data.
3. The marine disturbance estimation and observer design method using multi-source sensor fusion according to claim 1, wherein the conversion for wave disturbances in the "Northeast (NED) coordinate system conversion" in step S1, in particular, comprises the following sub-steps: S14, constructing a rotation matrix, namely constructing the rotation matrix fusing ship postures , Wherein Is a roll angle, Is a pitch angle, Is a course angle; s15, calculating a ship speed vector, namely converting the wave interference into the wave interference according to the real-time navigational speed of the ship And course angle Constructing a ship speed vector Is the ship speed vector ; S16, converting wave interference coordinates, namely converting wave interference under a ship body coordinate system By rotating the matrix Superimposed with the ship speed influence term, the wave interference is converted into wave interference under the North east ground coordinate system, and the formula is Where t is the time interval.
4. The method for designing an offshore interference estimation and observer using multi-source sensor fusion according to claim 1, wherein the step S2 of constructing a wave interference model and a water flow interference model specifically comprises the following sub-steps: s21, constructing a wave interference model based on wave height in the preprocessing data Cycle of Phase of Direction and direction of Constructing a periodic wave interference model, wherein the formula is as follows ; S22, constructing a water flow interference model based on the water flow speed in the preprocessing data Direction of water flow Constructing a constant-speed water flow interference model with the formula of ; S23, calculating a total interference initial model, namely vector superposition is carried out on the wave interference model and the water flow interference model to obtain the total interference initial model, wherein the formula is as follows 。
5. The method for designing an offshore interference estimation and observer using multi-source sensor fusion according to claim 1, wherein the step of converting into a state equation and an observation equation in step S3 comprises the following steps: s31, constructing a state equation by using the total interference initial model For state quantity, introducing control input ( Is the ship speed, Heading), zero-mean gaussian process noise (Covariance is Q) and a state equation is constructed, wherein the formula is Wherein A is a state transition matrix and B is an input matrix; S32, constructing an observation equation by using the sensor observation value For output quantity, zero-mean Gaussian observation noise is introduced (Covariance is ) And observe matrix Set as a unit matrix, and construct an observation equation with the formula of 。
6. The method for designing an offshore interference estimation and observer using multi-source sensor fusion according to claim 1, wherein the step S4 of predicting and updating the interference estimation value by using a kalman filter algorithm comprises the following steps: S41, initializing parameters, namely setting initial interference estimated values Initial covariance matrix ( Unit matrix) and determining the process noise covariance Q # ) Covariance R of observed noise ); S42, predicting interference state according to the current interference estimated value With control input Calculation of The predicted interference value at the moment is expressed as Simultaneously calculating a prediction covariance matrix, wherein the formula is ; S43, updating interference estimation, receiving Sensor observations at time of day First calculate Kalman gain Then based on Kalman gain, the predicted interference value is corrected to obtain an updated interference estimated value, the formula is And finally updating the covariance matrix according to the formula 。
7. The method of marine disturbance estimation and observer design using multisource sensor fusion according to claim 1, wherein the "generating a final control signal" in step S5 specifically comprises the sub-steps of: s51, calculating feedforward compensation signals based on the real-time interference estimated value obtained in the step S4 Combining the optimized gain matrix The feedforward compensation signal is calculated by the formula , wherein, Is a feedforward control compensation signal; is an estimated state of interference; s52, acquiring expected control input, namely calculating the expected control input according to the deviation between the target track of the ship and the current state through a PID (proportion integration differentiation) or LQR (Linear quick response) algorithm ; S53, synthesizing a final control signal, namely superposing the feedforward compensation signal and the expected control input to obtain the final control signal, wherein the formula is , wherein, Is a control input based on a ship dynamics model and target track planning.
8. The method of marine interference estimation and observer design using multi-source sensor fusion according to claim 1, further comprising the steps of: And S7, performing hardware deployment and algorithm execution, namely configuring at least one processor and at least one memory in communication connection with the processor, storing program instructions executable by the processor in the memory, and sequentially executing steps S1 to S6 when the processor invokes the program instructions to realize the estimation and observation of offshore interference, wherein the data acquisition of an anemograph can be increased in the step S1, and an interference model is expanded to be a wave + water flow + wind speed interference model in the step S2 so as to adapt to the dynamic positioning scene of an offshore platform or an intelligent ship.

Description

Marine interference estimation and observer design method using multi-source sensor fusion Technical Field The invention belongs to the field of offshore platform or intelligent ship control, and particularly relates to a design method of an offshore interference estimation and observer by using multi-source sensor fusion. Background The traditional ship control system mainly relies on basic sensors such as GPS and IMU to measure the position, speed and other states of the ship, wave and water flow sensor data are not fused, external interference such as water flow and wave has high uncertainty, accurate prediction cannot be performed through a conventional motion model, so that interference estimation errors are large, and navigation precision and stability are affected. In the existing multi-sensor application, the output frequency difference of each sensor is obvious (for example, the IMU is 10Hz, the GPS is 5Hz, the wave sensor is 10 min/time), time sequence deviation exists in data, the measured values of the wave sensor and the water flow sensor are based on a ship body coordinate system, are not unified with the north east coordinate system of the GPS and cannot be directly fused, and the interference estimation precision is further reduced. In addition, the existing interference estimation model does not combine the ship course and the ship speed, only calculates interference based on local coordinate data, and cannot be converted into real interference under an absolute coordinate system, so that the interference compensation direction and the amplitude deviation are caused, and the control system is difficult to effectively offset the environmental influence. The traditional ship control only adopts feedback control (such as PID), under the sea condition of 1-3 levels, the position deviation reaches 0.6-3.2m, the heading error reaches 1.5-6.8 degrees, the rudder angle change rate reaches 3.0-8.1 degrees/s, the error is larger under the complex sea condition, and the frequent action of the steering engine causes high energy consumption and quick equipment loss. There is therefore a need to solve the above problems using multi-source sensor fusion marine interference estimation and observer design methods. Disclosure of Invention An objective of the embodiments of the present invention is to provide a method for designing an offshore interference estimation and observer using multi-source sensor fusion, so as to solve the problems set forth in the background art. In order to achieve the above purpose, the present invention provides the following technical solutions: An offshore interference estimation and observer design method using multi-source sensor fusion, comprising the steps of: S1, sensor data acquisition and preprocessing, namely acquiring data of an Inertial Measurement Unit (IMU), a Global Positioning System (GPS), a wave sensor and a water flow sensor through a sensor data acquisition and preprocessing module, respectively performing time alignment and North East (NED) coordinate system conversion on the acquired data to obtain preprocessed data, and outputting the preprocessed data to an interference estimation model construction module; S2, constructing an interference estimation model, namely respectively constructing a wave interference model and a water flow interference model based on the received pretreatment data by an interference estimation model construction module, superposing the wave interference model and the water flow interference model to obtain a total interference initial model, and outputting the total interference initial model to a state space model module; s3, converting the state space model, namely converting the total interference initial model into a state equation describing interference dynamic change by a state space model module, and outputting the state equation and the observation equation to an interference observer design module by an observation equation describing the relation between sensor observation data and interference state; S4, estimating an interference observer, namely, based on a state equation and an observation equation, adopting a Kalman filtering algorithm by an interference observer design module, firstly predicting an interference state and a covariance matrix according to the state equation, then updating an interference estimation value by combining sensor observation data, and outputting the real-time interference estimation value to a feedforward controller design module; S5, generating a feedforward control signal, namely receiving a real-time interference estimated value by a feedforward controller design module, generating a final control signal by combining an expected control input obtained based on track planning, a target course and a speed, and outputting the final control signal to a control execution module; and S6, controlling execution, namely adjusting the rudder angle and the accelerator of the ship by the control e