A surface ship multi-source speed information fusion method

Abstract

The application relates to a water surface ship multi-source speed information fusion method, which is based on multi-source speed error characteristics to establish a multi-source speed error mathematical model, utilizes high complementarity of speed error, and realizes real-time calculation of multi-speed source fusion speed based on Kalman filtering technology to obtain high-quality ground speed information formed by multi-speed fusion. The calculated speed information is provided for the ship, and the speed performance of the comprehensive navigation system can be effectively improved; after the speed information is used for a damping network of an inertial navigation system, the navigation information precision of the inertial navigation system is obviously improved, and the navigation information level of the comprehensive navigation system is further improved. Meanwhile, the application does not change the hardware structure of the comprehensive navigation system, can be widely applied to the comprehensive navigation system of the water surface ship, and has good engineering application prospect and popularization value.

Description

Technical Field

[0001] This invention belongs to the field of marine multi-source speed information technology, and in particular, a method for fusing multi-source speed information of surface ships. Background Technology

[0002] With the rapid development of my country's shipbuilding industry, the speed measurement equipment on surface ships has become increasingly diverse. Currently, commonly used speed measurement equipment on surface ships mainly includes electromagnetic logs, Doppler logs, and satellite navigation equipment. These speed measurement devices operate on different principles and have different speed error characteristics, but these characteristics are highly complementary. In current shipboard applications, typically only one type of speed measurement equipment is used. However, if the strong complementarity of different external speed error characteristics could be utilized to overcome the shortcomings of a single speed measurement device, and two or three types of external speed information could be fused to form higher-quality speed information, a higher quality result could be achieved. Summary of the Invention

[0003] The purpose of this invention is to overcome the shortcomings of the prior art and propose a multi-source velocity information fusion method for surface ships, which can overcome the shortcomings of a single velocity measuring device for surface ships and make full use of the strong complementary characteristics of multi-source velocity error.

[0004] The technical problem solved by this invention is achieved through the following technical solution:

[0005] A method for fusing multi-source velocity information of surface ships includes the following steps:

[0006] Step 1: Construct a mathematical model for multi-source velocity information error;

[0007] Step 2: Construct a multi-source velocity information fusion algorithm based on Kalman filtering;

[0008] Step 3: Based on the current situation of the ship, use the multi-source speed information error mathematical model constructed in Step 1 to calculate the speed measurement error using the fusion algorithm constructed in Step 2. At the same time, the speed measurement error is compensated into the speed of the electromagnetic log.

[0009] Furthermore, the multi-source velocity information error mathematical model in step 1 includes: electromagnetic log velocity error model, Doppler log ground velocity error model, and satellite navigation velocity error model.

[0010] Furthermore, the specific implementation method of the electromagnetic speed error model is as follows:

[0011]

[0012] Among them, X E Let X be the state vector of the electromagnetic log speed model. E =[δV exδV ey δK e V sx V sy ] T ;F E W is the state transition matrix for the speed model of the electromagnetic log. E The noise vector of the electromagnetic odometer speed model;

[0013]

[0014]

[0015] Where, δV ex For the lateral velocity error of the electromagnetic log; δV ey For the longitudinal speed error of the electromagnetic log; V sx V represents the lateral velocity of the ocean current. sy δK represents the longitudinal velocity of the ocean current. e For the electromagnetic odometer scale coefficient error; V x V is the lateral velocity of the ship. y ω represents the longitudinal velocity of the ship. ex For the lateral velocity noise of the electromagnetic log; τ sx The transverse velocity correlation time for a first-order Markov process of ocean currents; τ sy For the longitudinal velocity correlation time of the first-order Markov process of ocean currents; ω sx For the transverse noise of the ocean current velocity model; ω sy This represents the longitudinal noise of the ocean current velocity model.

[0016] Furthermore, the specific implementation method of the Doppler log's ground velocity error model is as follows:

[0017]

[0018] Among them, X D Let X be the state vector of the Doppler log velocity model. D =[δV dx δV dy δK d ] T ;F D W is the state transition matrix for the Doppler log velocity model. D The noise vector of the Doppler log velocity model;

[0019]

[0020] Where, δV dx The lateral ground velocity error of the Doppler log; δV dy For the longitudinal ground velocity error of the Doppler log; δK dFor the Doppler log scale coefficient error; ω dx For the lateral velocity noise of the Doppler log; ω dy α represents the longitudinal velocity noise of the Doppler log; α represents the installation angle error of the Doppler log.

[0021] Furthermore, the specific implementation method of the satellite guidance velocity error model is as follows:

[0022]

[0023] Among them, X W X is the state vector of the satellite guidance velocity model; W =[δV WE +δV WN ] T ;F W W is the state transition matrix for the satellite guidance velocity model. W This represents the noise vector of the satellite navigation velocity model.

[0024]

[0025] Where, ω WE For satellite guidance noise; ω WN The noise is due to the high speed of the satellite.

[0026] Furthermore, the specific implementation method of step 2 is as follows:

[0027]

[0028] in, This is the estimated value of the state vector; Z is the estimated value for the state in one step; k For the observation sequence; Φ k,k-1 K is the state transition matrix; k H is the optimal filter gain matrix; k P is the observation matrix; k Let P be the error variance matrix; k,k-1 Γ is the variance matrix of the prediction error in one step; k,k-1 R is the noise input matrix; k For the system observation noise sequence V k The variance matrix of Q; k The system process noise sequence W k The variance matrix.

[0029] Furthermore, step 3 includes the following steps:

[0030] Step 3.1: Determine whether the ship can obtain the electromagnetic log speed, Doppler ground speed, and satellite navigation speed. If the ship can obtain the electromagnetic log speed and satellite navigation speed, but cannot obtain the Doppler ground speed, proceed to step 3.2, and use the fused speed from step 3.2 as the non-inertial speed of the integrated navigation system for the ship's use, and at the same time as the damping speed of the inertial navigation system to improve the overall navigation accuracy of the inertial navigation system and even the integrated navigation system.

[0031] If the ship can obtain the electromagnetic log speed, Doppler ground speed and satellite navigation speed, proceed to step 3.3, and use the fused speed in step 3.3 as the non-inertial speed of the integrated navigation system for the ship's use, and at the same time as the damping speed of the inertial navigation system, to improve the overall navigation accuracy of the inertial navigation system and even the integrated navigation system.

[0032] Step 3.2: Based on the multi-source velocity information error mathematical model in Step 1, which includes the electromagnetic log velocity error model and the satellite navigation velocity error model, the ocean current, the electromagnetic log's own velocity measurement error, and the satellite navigation velocity error are used as state variables, and the difference between the electromagnetic log velocity and the satellite navigation velocity is used as the observation vector. These are substituted into the multi-source velocity information fusion algorithm based on Kalman filtering constructed in Step 2 to estimate the ocean current and the electromagnetic log's own velocity measurement error in real time and compensate it into the electromagnetic log velocity. At the same time, the absolute velocity formed by the fusion of the electromagnetic log velocity and the satellite navigation velocity is obtained as the fused velocity I.

[0033] Step 3.3: Based on the multi-source velocity information error mathematical model in Step 1, which includes the electromagnetic log velocity error model, the Doppler log ground velocity error model, and the satellite navigation velocity error model, the ocean current, the electromagnetic log's own velocity measurement error, the satellite navigation velocity error, and the Doppler log ground velocity error are used as state variables. The differences between the electromagnetic log velocity and the satellite navigation velocity, the differences between the satellite navigation velocity and the Doppler log ground velocity, and the differences between the electromagnetic log velocity and the Doppler log ground velocity are used as observation vectors. These are substituted into the multi-source velocity information fusion algorithm based on Kalman filtering constructed in Step 2 to estimate the ocean current and the electromagnetic log's own velocity measurement error in real time and compensate for it in the electromagnetic log velocity. At the same time, the absolute velocity formed by the fusion of the electromagnetic log velocity, the satellite navigation velocity, and the Doppler log ground velocity is obtained as the fused velocity II.

[0034] The advantages and positive effects of this invention are:

[0035] This invention establishes a multi-source velocity error mathematical model based on the characteristics of multi-source velocity errors. Utilizing the high complementarity of these velocity errors, and based on Kalman filtering technology, it calculates the fused velocity from multiple velocity sources in real time, obtaining high-quality ground-to-surface velocity information. The calculated velocity information is provided to ships, effectively improving the velocity performance of the integrated navigation system. When this velocity information is used in the damping network of the inertial navigation system, it significantly improves the accuracy of the inertial navigation system's navigation information, thereby enhancing the overall navigation information level of the integrated navigation system. Furthermore, this invention does not alter the hardware structure of the integrated navigation system and can be widely applied to surface ship integrated navigation systems, demonstrating excellent engineering application prospects and promotional value. Attached Figure Description

[0036] Figure 1 This is a comparison diagram of the single external speed and the reference speed of the present invention;

[0037] Figure 2 This is a comparison chart of the fusion speed I of this invention and the reference speed;

[0038] Figure 3 This is a comparison chart of the fusion speed II of this invention and the reference speed. Detailed Implementation

[0039] The present invention will be further described in detail below with reference to the accompanying drawings.

[0040] A method for fusing multi-source velocity information of surface ships, characterized by the following steps:

[0041] Step 1: Construct a mathematical model for multi-source velocity information errors. The mathematical model for multi-source velocity information errors includes: an electromagnetic log velocity error model, a Doppler log ground velocity error model, and a satellite navigation velocity error model.

[0042] Electromagnetic speed log speed error model:

[0043] Electromagnetic logs measure velocity relative to water; therefore, the absolute error of an electromagnetic log's velocity measurement includes not only its own relative velocity measurement error and scale coefficient error, but also the ocean current velocity. Since ocean current velocity is time- and area-dependent and exhibits random variations, a first-order Markov process ocean current velocity model is established:

[0044]

[0045] Among them, V sx V represents the lateral velocity of the ocean current. sy τ is the longitudinal velocity of the ocean current. sx The transverse velocity correlation time for a first-order Markov process of ocean currents; τ sy For the longitudinal velocity correlation time of the first-order Markov process of ocean currents; ω sx For the transverse noise of the ocean current velocity model; ωsy This represents the longitudinal noise of the ocean current velocity model.

[0046] The mathematical model for the speed error of an electromagnetic log consists of three parts: the ocean current model, the electromagnetic log's relative speed measurement error model, and the electromagnetic log's scale coefficient error model. Based on this, the mathematical model for the speed error of the electromagnetic log is established as follows:

[0047]

[0048] Where, δV ex For the lateral velocity error of the electromagnetic log; δV ey For the longitudinal speed error of the electromagnetic log; V sx V represents the lateral velocity of the ocean current. sy δK represents the longitudinal velocity of the ocean current. e For the electromagnetic odometer scale coefficient error; V x V is the lateral velocity of the ship. y ω represents the longitudinal velocity of the ship. ex This refers to the lateral velocity noise of the electromagnetic log.

[0049] The first-order Markov process ocean current velocity model and the electromagnetic log velocity error mathematical model are converted into the following state equations:

[0050]

[0051] Among them, X E Let X be the state vector of the electromagnetic log speed model. E =[δV ex δV ey δK e V sx V sy ] T ;F E W is the state transition matrix for the speed model of the electromagnetic log. E This is the noise vector of the electromagnetic log speed model.

[0052] The ground velocity error of a Doppler log consists of three parts: velocity measurement error noise, scale coefficient error, and installation angle error. The ground velocity error model of a Doppler log is as follows:

[0053]

[0054] Where, δV dx The lateral ground velocity error of the Doppler log; δV dy For the longitudinal ground velocity error of the Doppler log; δK d For the Doppler log scale coefficient error; ω dx For the lateral velocity noise of the Doppler log; ω dyα represents the longitudinal velocity noise of the Doppler log; α represents the installation angle error of the Doppler log.

[0055] The Doppler log's ground velocity error model is converted into a state equation as follows:

[0056]

[0057] Among them, X D Let X be the state vector of the Doppler log velocity model. D =[δV dx δV dy δK d ] T ;F D W is the state transition matrix for the Doppler log velocity model. D The noise vector of the Doppler log velocity model;

[0058] The satellite navigation velocity error is in the form of noise, and its mathematical model is as follows:

[0059]

[0060] Where, ω WE For satellite guidance noise; ω WN The noise is due to the high speed of the satellite.

[0061] The mathematical model of satellite guidance velocity error is transformed into a state equation as follows:

[0062]

[0063] Among them, X W X is the state vector of the satellite guidance velocity model; W =[δV WE +δV WN ] T ;F W W is the state transition matrix for the satellite guidance velocity model. W This is the noise vector of the satellite navigation velocity model.

[0064] Step 2: Construct a multi-source velocity information fusion algorithm based on Kalman filtering. Kalman filtering is one of the main techniques for modern multi-sensor information fusion. Kalman filtering estimates all signals to be processed based on the system equations and observation equations. This invention uses the Kalman filtering method to design an external velocity information fusion processing algorithm. Assume the state equations and observation equations of the stochastic continuous system are as follows:

[0065]

[0066] Where F is the state transition matrix of the continuous system; X is the state vector of the continuous system; and W is the noise vector of the system's stochastic process.

[0067] Z = HX + V

[0068] Where Z is the observation vector of the continuous system; H is the observation matrix; X is the state vector of the continuous system; and V is the random observation noise vector of the system.

[0069] To facilitate engineering applications, the continuous state equations are discretized to obtain the state equations and observation equations for the stochastic discrete system, as follows:

[0070] X k =Φ k,k-1 X k-1 +Γ k,k-1 W k-1

[0071] Among them, X k W represents the state vector of a discrete system. k For the system's random process noise sequence; Φ k,k-1 Γ is the state transition matrix; k,k-1 This is the noise input matrix.

[0072] Z k =H k X k +V k

[0073] Among them, Z k For observation sequence; H k X is the observation matrix; k V is the state vector of the discrete system; k It is a noise sequence.

[0074] The fundamental equations for Kalman filtering of discrete systems are:

[0075]

[0076] in, This is the estimated value of the state vector; Z is the estimated value for the state in one step; k For observation sequence;

[0077] Φ k,k-1 K is the state transition matrix; k H is the optimal filter gain matrix; k P is the observation matrix; k Let P be the error variance matrix; k,k-1 Γ is the variance matrix of the prediction error in one step; k,k-1 R is the noise input matrix; k For the system observation noise sequence V k The variance matrix of Q; k The system process noise sequence Wk The variance matrix.

[0078] Step 3: Based on the current situation of the ship, use the multi-source speed information error mathematical model constructed in Step 1 to calculate the speed measurement error using the fusion algorithm constructed in Step 2. At the same time, the speed measurement error is compensated into the speed of the electromagnetic log.

[0079] Step 3.1: Determine whether the ship can obtain the electromagnetic log speed, Doppler ground speed, and satellite navigation speed. If the ship can obtain the electromagnetic log speed and satellite navigation speed, but cannot obtain the Doppler ground speed, proceed to step 3.2, and use the fused speed from step 3.2 as the non-inertial speed of the integrated navigation system for the ship's use, and at the same time as the damping speed of the inertial navigation system to improve the overall navigation accuracy of the inertial navigation system and even the integrated navigation system.

[0080] If the ship can obtain the electromagnetic log speed, Doppler ground speed and satellite navigation speed, proceed to step 3.3, and use the fused speed in step 3.3 as the non-inertial speed of the integrated navigation system for the ship's use, while also using it as the damping speed of the inertial navigation system to improve the overall navigation accuracy of the inertial navigation system and even the integrated navigation system.

[0081] Step 3.2: Based on the multi-source velocity information error mathematical model in Step 1, which includes the electromagnetic log velocity error model and the satellite navigation velocity error model, the ocean current, the electromagnetic log's own velocity measurement error, and the satellite navigation velocity error are used as state variables. The difference between the electromagnetic log velocity and the satellite navigation velocity is used as the observation vector. These are substituted into the multi-source velocity information fusion algorithm based on Kalman filtering constructed in Step 2 to estimate the ocean current and the electromagnetic log's own velocity measurement error in real time and compensate for it in the electromagnetic log velocity. At the same time, the absolute velocity formed by the fusion of the electromagnetic log velocity and the satellite navigation velocity is obtained as the fused velocity I. This scheme is called Fusion Scheme I.

[0082] Step 3.3: Based on the multi-source velocity information error mathematical model in Step 1, which includes the electromagnetic log velocity error model, the Doppler log ground velocity error model, and the satellite navigation velocity error model, the ocean current, the electromagnetic log's own velocity measurement error, the satellite navigation velocity error, and the Doppler log ground velocity error are used as state variables. The differences between the electromagnetic log velocity and the satellite navigation velocity, the differences between the satellite navigation velocity and the Doppler log ground velocity, and the differences between the electromagnetic log velocity and the Doppler log ground velocity are used as observation vectors. These are substituted into the multi-source velocity information fusion algorithm based on Kalman filtering constructed in Step 2 to estimate the ocean current and the electromagnetic log's own velocity measurement error in real time and compensate for it in the electromagnetic log velocity. At the same time, the absolute velocity formed by the fusion of the electromagnetic log velocity, the satellite navigation velocity, and the Doppler log ground velocity is obtained as the fused velocity II. This scheme is called Fusion Scheme II.

[0083] According to the present invention, a method for fusing multi-source velocity information of surface ships has been tested on a certain ship, and the correctness of the present invention has been proven.

[0084] Among them, the appendix Figure 1 A comparison chart of a single external speed and a reference speed is attached. Figure 2 This is a comparison chart of fusion speed I and the baseline speed. (Attached) Figure 3 This is a comparison chart of fusion speed II and the reference speed. Based on the above chart and the inherent characteristics of a single speed, the present invention can be understood as follows:

[0085] Electromagnetic speed logs have good speed stability, but due to their speed measurement principle, they have large speed errors during large ship maneuvers (such as turning).

[0086] The stability of the Doppler log to ground velocity is lower than that of fusion velocity I and fusion velocity II, and the Doppler ground velocity is affected by water depth and can only be obtained within a certain range.

[0087] The average error of satellite-guided velocity measurement is small, but the overall measurement noise is large and the information stability is poor.

[0088] The accuracy and stability of the velocity information of Fusion Speed ​​I and Fusion Speed ​​II have always remained high, and Fusion Speed ​​II has better overall quality than Fusion Speed ​​I due to the increase in observation information.

[0089] Based on the above comparison, this invention can effectively improve the accuracy and quality of navigation speed information for surface ships. On the one hand, it provides ships with more stable and high-precision speed information, enhancing the speed information output level of the integrated navigation system. On the other hand, marine inertial navigation systems typically require external speed information to operate in a damped mode. Step and slow variations in external speed information can disturb the navigation information of the inertial navigation system, especially the horizontal speed information. Using fused speed instead of a single external speed for inertial navigation system damping, due to its high precision and stability, will significantly improve the navigation accuracy of the inertial navigation system, thereby improving the overall navigation accuracy of the integrated navigation system. Therefore, the realization of this invention has significant military and civilian value.

[0090] It should be emphasized that the embodiments described in this invention are illustrative rather than limiting. Therefore, this invention includes, but is not limited to, the embodiments described in the specific implementation. Any other implementations derived by those skilled in the art based on the technical solutions of this invention are also within the scope of protection of this invention.

Claims

1. A method for fusing multi-source velocity information of surface ships, characterized in that: Includes the following steps: Step 1: Construct a mathematical model for multi-source velocity information error; The mathematical models for multi-source velocity information errors include: electromagnetic log velocity error model, Doppler log ground velocity error model, and satellite navigation velocity error model; The specific implementation method of the electromagnetic speed error model is as follows: in, This represents the state vector of the electromagnetic odometer speed error model. ; This is the state transition matrix for the speed error model of an electromagnetic log. This represents the noise vector of the electromagnetic log speed error model. in, This refers to the lateral velocity error of the electromagnetic log. This refers to the longitudinal speed error of the electromagnetic log. This refers to the lateral velocity of the ocean current. The longitudinal velocity of the ocean current; This refers to the calibration coefficient error of the electromagnetic log. The lateral velocity of the ship; The longitudinal velocity of the ship; This refers to the lateral velocity noise of the electromagnetic log. For the transverse velocity correlation time of the first-order Markov process of ocean currents; For the longitudinal velocity correlation time of the first-order Markov process of ocean currents; Lateral noise in the ocean current velocity model; For longitudinal noise in the ocean current velocity model; The specific implementation method of the Doppler log's ground velocity error model is as follows: in, Let the state vector of the Doppler log speed error model be... ; The state transition matrix for the Doppler log velocity error model; The noise vector of the Doppler log speed error model; in, This refers to the lateral ground velocity error of the Doppler log. This refers to the longitudinal ground velocity error of the Doppler log. This refers to the Doppler log scale coefficient error. The noise is the lateral velocity of the Doppler log. This refers to the longitudinal velocity noise of the Doppler log. To correct the angle error of the Doppler log; The specific implementation method of the satellite guidance velocity error model is as follows: in, This represents the state vector of the satellite navigation velocity error model. ; This is the state transition matrix for the satellite navigation velocity error model; This represents the noise vector of the satellite navigation velocity error model. in, For satellite guidance, the noise level is high. For satellite guidance, high-speed noise; Step 2: Construct a multi-source velocity information fusion algorithm based on Kalman filtering; in, This is the estimated value of the state vector; The state is predicted and estimated in one step; For observation sequence; This is the state transition matrix; This is the optimal filter gain matrix; The observation matrix; Let Variance be the error matrix; The variance matrix of the prediction error in one step; The noise input matrix; For system observation noise sequence The variance matrix; The system process noise sequence at time k-1 The variance matrix; Step 3: Based on the current situation of the ship, use the multi-source speed information error mathematical model constructed in Step 1 to calculate the speed measurement error in the fusion algorithm constructed in Step 2, and then compensate the speed measurement error into the speed of the electromagnetic log. Step 3.1: Determine whether the ship can obtain the electromagnetic log speed, Doppler ground speed, and satellite navigation speed. If the ship can obtain the electromagnetic log speed and satellite navigation speed, but cannot obtain the Doppler ground speed, proceed to step 3.2, and use the fused speed in step 3.2 as the non-inertial speed of the integrated navigation system for the ship's use, and at the same time as the damping speed of the inertial navigation system to improve the overall navigation accuracy of the inertial navigation system and even the integrated navigation system. If the ship can obtain the electromagnetic log speed, Doppler ground speed and satellite navigation speed, proceed to step 3.3, and use the fused speed in step 3.3 as the non-inertial speed of the integrated navigation system for the ship's use, while also using it as the damping speed of the inertial navigation system to improve the overall navigation accuracy of the inertial navigation system and even the integrated navigation system. Step 3.2: Based on the multi-source velocity information error mathematical model in Step 1, which includes the electromagnetic log velocity error model and the satellite navigation velocity error model, the ocean current, the electromagnetic log's own velocity measurement error, and the satellite navigation velocity error are used as state variables, and the difference between the electromagnetic log velocity and the satellite navigation velocity is used as the observation vector. These are substituted into the multi-source velocity information fusion algorithm based on Kalman filtering constructed in Step 2 to estimate the ocean current and the electromagnetic log's own velocity measurement error in real time and compensate it into the electromagnetic log velocity. At the same time, the absolute velocity formed by the fusion of the electromagnetic log velocity and the satellite navigation velocity is obtained as the fused velocity I. Step 3.3: Based on the multi-source velocity information error mathematical model in Step 1, which includes the electromagnetic log velocity error model, the Doppler log ground velocity error model, and the satellite navigation velocity error model, the ocean current, the electromagnetic log's own velocity measurement error, the satellite navigation velocity error, and the Doppler log ground velocity error are used as state variables. The differences between the electromagnetic log velocity and the satellite navigation velocity, the differences between the satellite navigation velocity and the Doppler log ground velocity, and the differences between the electromagnetic log velocity and the Doppler log ground velocity are used as observation vectors. These are substituted into the multi-source velocity information fusion algorithm based on Kalman filtering constructed in Step 2 to estimate the ocean current and the electromagnetic log's own velocity measurement error in real time and compensate for it in the electromagnetic log velocity. At the same time, the absolute velocity formed by the fusion of the electromagnetic log velocity, the satellite navigation velocity, and the Doppler log ground velocity is obtained as the fused velocity II.

Images