Unmanned ship positioning method based on multi-sensor data fusion

Abstract

An unmanned ship positioning method based on multi-sensor data fusion, the steps of which are as follows: first, pre-process data collected by an unmanned ship multi-sensor positioning system, then perform a reliability judgment on the unmanned ship multi-sensor positioning data through a reliability distance test, and assign a corresponding reliability factor to the tested positioning data, perform a test compensation for faulty data using a consistency test and variance weighting, and perform a filtering process on each sensor positioning data after testing and weighting based on basic particle filtering to realize data augmentation, and finally perform a fusion filtering output process on the unmanned ship multi-sensor positioning data using a new threshold hierarchical particle filtering algorithm to obtain accurate unmanned ship navigation trajectory positioning information. The method improves the fault tolerance performance and algorithm relevance of the unmanned ship multi-sensor positioning system, ensures the reliability of the sensor positioning data, and achieves the purpose of accurately positioning the unmanned ship navigation trajectory.

Description

[Technical field]

[0001] The present invention belongs to the technical field of positioning and navigation, and relates to a fusion processing technology based on multi-sensor information collection, and particularly to an unmanned ship positioning method based on multi-sensor data fusion. [Background technology]

[0002] The development and application of intelligent unmanned ship positioning systems supported by modern sensor information technology can not only promote the development of surface unmanned ships in areas such as coastal patrol, water quality monitoring, and aquaculture, but also greatly reduce the work intensity of workers and improve work efficiency.

[0003] In order to realize the efficient and accurate work of unmanned ships, we must first rely on whether the obtained sensor positioning information can accurately sense the position of the surface unmanned ship, that is, whether the unmanned ship can use the obtained positioning information to ensure its work efficiency and accuracy. With the development of automatic navigation technology for surface unmanned ships, the requirements for the positioning accuracy and stability of unmanned ships are increasing. Due to the influence of the complex working environment on the water surface and shore, sensor information may be lost and the positioning may be inaccurate. Therefore, there are great disadvantages when using a single sensor, and in the case of multi-sensor data fusion, the above disadvantages can be effectively overcome by using their advantages in a complementary manner, which has become the development trend of the combined positioning system. Therefore, the research and application of the multi-sensor data fusion algorithm of the unmanned ship positioning system is developed, which is helpful to improve the fault tolerance performance of the algorithm, and further realizes the highly efficient filtering processing of the unmanned ship position data by the multi-sensor data fusion algorithm, and finally achieves the accurate positioning of the unmanned ship position and the improvement of production work efficiency. Summary of the Invention [Problem to be solved by the invention]

[0004] The objective of the present invention is to provide an unmanned ship positioning method based on multi-sensor data fusion for multi-sensor data fusion positioning of unmanned ship navigation trajectory, so as to overcome the major disadvantages of the prior art due to the use of a single sensor. [Means for solving the problem]

[0005] In order to achieve the above object, the present invention is realized using the following technical solutions.

[0006] An unmanned ship positioning method based on multi-sensor data fusion, which first pre-processes the positioning data collected by the unmanned ship multi-sensor positioning system, then performs a confidence distance judgment on the unmanned ship positioning data and assigns a corresponding confidence factor, simultaneously performs fault checking and weight compensation on the positioning data, and performs filtering augmentation on the positioning data, and finally uses a data fusion algorithm to perform multi-sensor data fusion filtering output, thereby realizing the positioning of the unmanned ship navigation trajectory, the specific steps of which are as follows: Step 1. Data preprocessing, This paper preprocesses unmanned ship positioning data collected by an unmanned ship multi-sensor positioning system.

[0007] Step 2. Data reliability determination and allocation, The data reliability determination and assignment includes performing reliability determination and reliability factor assignment based on data collected by multiple sensors.

[0008] Step 3. Data failure inspection compensation, The data fault checking compensation includes performing a weighting compensation process on the inconsistency data collected by the positioning system by performing a consistency check on the multi-sensor data.

[0009] Step 4. Data Augmentation, The data augmentation includes performing filtering processes on the pre-processed, inspected and compensated multi-sensor data respectively based on a basic particle filtering algorithm.

[0010] Step 5. Data fusion The data fusion includes designing a novel threshold hierarchical particle filtering algorithm by constructing a Gaussian mixture model and setting a hierarchical sampling rate capacity associated with an adaptive threshold and a confidence factor, and further performing fusion filtering on the unmanned ship navigation trajectory positioning data to output unmanned ship navigation trajectory positioning information.

[0011] In addition, the specific content and method steps of the data pre-processing described in step 1 of pre-processing the position data collected by the unmanned ship multi-sensor positioning system are as follows: A) Based on the longitude and latitude information of the unmanned ship's navigation trajectory collected by the positioning system sensor, the temporal and spatial standards will be unified.

[0012] B) According to the river route along which the unmanned ship navigates, a communication environment that satisfies the effective connection of multiple nodes of the positioning system is constructed, so that the blind node located on the unmanned ship is within a network of four signal receiving nodes with known transmission signal strength and location coordinates, thereby calculating and collecting the location coordinates of the unmanned ship as it navigates the river route.

[0013] C) Transform and unify the multi-sensor positioning system coordinates based on the Gauss-Krüger projection principle.

[0014] Furthermore, the data reliability judgment and allocation in step 2 includes making a reliability judgment and a reliability factor allocation according to the data collected by the multi-sensor, and the specific content and method steps of the data reliability judgment are as follows: 1) Measurement model of measurement data of unmanned ship positioning system p i According to (x), the confidence distance d between the sensor data i Calculate the p i(x), d i The calculation formula is as follows:

number

number

number

[0015] 2) The d obtained in step 1) according to the Gaussian probability model i Let p be the probability semantic metric between the sensor measurement data. r (Z i ) and set the sensor support reliability level, which determines the reliability of different sensor information. r (Z i The formula for calculating

number

[0016] 3) According to the magnitude of the probability semantic information metric value between the measurement data, the determined positioning data is divided into different confidence intervals so that the reliable data approaches the high confidence region, and the set confidence level ε is determined by the sensor measurement variance average value

number

number

[0017] Furthermore, in the data reliability judgment and assignment described in step 2, the specific content and method steps of assigning the data reliability factor are as follows: (A) Dynamic support factor β of measurement information collected by multiple sensors at time k i (k) and the Gaussian probability measurement model p i (k) norm equation is constructed, and the β i The formula for (k) is as follows:

number

[0018] (B) The multi-sensor dynamic support factor β calculated in step (A). i Based on (k), the measurement error of the system w i Calculate the w i The calculation formula is as follows: w i =z i -Aβ i (k) In the formula, A is the state transition matrix, and z i are the measurements of the unmanned ship multi-sensor positioning system, for i = 1, 2, ..., n.

[0019] (C) System measurement error w i The variance of

number

number

[0020] According to the degree of reliability of the calculated unmanned ship multi-sensor measurement data, a corresponding reliability factor is assigned thereto, and the reliability factor is associated with a novel threshold hierarchical particle filtering algorithm, such that the higher the reliability of the measurement data, the higher the corresponding fusion bias will be, thereby improving the processing accuracy of the unmanned ship multi-sensor data fusion algorithm.

[0021] In addition, the data fault checking compensation described in step 3 includes performing a consistency check on the multi-sensor data, and performing a weighting compensation process on the inconsistent fault data collected by the positioning system. The specific content and method of the consistency check include the following steps: (1) The sensor measurement data obtained by the unmanned ship positioning system is arithmetically averaged to obtain the arithmetic mean

number

[0022] In addition, the method of performing weighted compensation processing on the mismatched fault data in the data fault inspection compensation described in step 3 specifically adopts the following steps: (I) Arithmetic mean value of the unmanned ship multi-sensor measurement data

number

number

number

number

number

number

[0023] Furthermore, the data augmentation in step 4 includes respectively performing filtering processes on the pre-processed, inspected and compensated multi-sensor data according to a basic particle filtering algorithm, and the specific content and method steps of the data augmentation are as follows: (i) Establish a basic particle filtering model of the unmanned ship multi-sensor positioning system, substitute the unmanned ship position information after consistency checking and variance weighting into the system model, and describe the state and measurement model of the sensor system as follows:

number

number

number

number

number

number

[0024] Particle propagation model for state transition of unmanned ship multi-sensor positioning system

number

[0025] In addition, the data fusion described in step 5 includes: designing a new threshold hierarchical particle filtering algorithm by constructing a Gaussian mixture model, and setting a hierarchical sampling proportion capacity associated with an adaptive threshold and a confidence factor; and performing fusion filtering on the unmanned ship navigation trajectory positioning data to output unmanned ship navigation trajectory positioning information. The specific content and method steps of constructing the Gaussian mixture model are as follows: (a) A multi-sensor dataset processed with a basic particle filtering algorithm and the Sigma point set

number

number

number

number

number

number

number

number

number

[0026] First, the posterior probability density function representing the state is represented by a continuous probability density function constructed with discrete particles, and the constructed continuous probability density function is approximated to a Gaussian mixture distribution. On top of that, particle samples are extracted instead of resampling, thereby maintaining particle diversity, avoiding sample deficiency, and improving filtering fusion accuracy.

[0027] Furthermore, the specific content and method steps of setting the adaptive threshold in the data fusion described in step 5 are as follows: a) After selecting the importance sampling process, the particle x with the largest weight in the discrete particle sample set is c is the clustering center, and the martensite distance D between other particles i and i Calculate the above D i The calculation formula is as follows:

number

number

number

number

number

[0028] The higher the probability density of a particle, the higher the D calculated in step a). i becomes smaller. The adaptive threshold T cSince is directly proportional to the particle probability density covariance, in step e) D i T compared to c D i Since the size can be adjusted according to the size of the Gaussian mixture, there is no need to constantly adjust the threshold to integrate similar units when clustering the Gaussian mixture, which reduces the number of clustering times for other particle samples and improves clustering efficiency. In addition, since the constructed adaptive threshold is related to the particle probability mass, when integrating similar units for a particle set, particles with similar weight magnitudes are clustered into one component unit based on the threshold size to ensure that the weight difference value between particle sets in the same component unit is minimum, and the weighted point set mixture Gaussian distribution constructed in step (e) is used to integrate similar units and then approximate them with the posterior probability density function, thereby improving the sampling accuracy in the resampling process.

[0029] In addition, in the data fusion described in step 5, the specific content and method steps of relating the reliability factor to a hierarchical sampling proportion capacity, and further performing fusion filtering on the unmanned ship navigation trajectory positioning data, and outputting the unmanned ship navigation trajectory positioning information are as follows: (1) Based on the hierarchy theory, each probability density function is a continuous probability density function of p(x).

number

[0030] (2) Each a , l b , l c Let the particle number fractional volumes of the layers be N / 4, N / 3, and N / 3.

[0031] (3) The reliability factor is substituted into the weight optimization combination and calculated.

[0032] (4)l b , l c The weights in the layer are the average

number

number

number

[0033] (6) The data fusion sampling results obtained in step (5) are output in the form of a log file from the blind node installed on the unmanned vessel.

[0034] (7) The PC-side coordinator node placed on the riverbed networks with the unmanned ship blind node to obtain the unmanned ship position information output by the blind node in step (7) in real time, thereby realizing positioning of the unmanned ship's navigation trajectory.

[0035] The sampling samples of the continuous probability density function of the weighted point set are stratified, and the proportion capacity of each sampling layer is set to ensure that the allocation of the number of sampling particles in the layer is reasonable; and a Sampling the layer set and l b , l cAn optimized combination is performed on the weights of particles in the layer set to increase their probability mass, and the reliability factor of the measurement data collected by the unmanned ship multi-sensor positioning system is associated with the hierarchical sampling in the multi-sensor data algorithm. When data fusion is performed on the unmanned ship multi-sensor measurement data using the new threshold hierarchical particle filtering algorithm, the sensor measurement values ​​with a large reliability factor are preferentially sampled and fused. Effect of the Invention

[0036] The present invention has the following advantages and beneficial effects.

[0037] (1) In the present invention, information is supplemented using positioning data collected by the unmanned ship's multi-sensor, effectively overcoming the problem of sensor signal loss due to environmental signal blockage, and ensuring the validity of the data collected by the sensors.

[0038] (2) In the present invention, a consistency check is performed on the unmanned ship position information collected by the multi-sensor positioning system, and weighted compensation is performed on the fault data, thereby improving the fault tolerance performance of the data fusion algorithm and ensuring the reliability of the particle set samples sampled by the basic particle filtering algorithm.

[0039] (3) Solve the integral operation in Bayesian estimation according to the Monte Carlo method adopted by basic particle filtering, eliminate environmental noise interference, ensure the processing accuracy of the sampling sample data set, and further improve the reliability of the unmanned ship multi-sensor measurement data.

[0040] (4) In the present invention, in the importance sampling process of the data fusion algorithm, the current measurement information is fused into the particle set proposal distribution, so that the proposal distribution is closer to the actual posterior probability density, and the algorithm estimation performance is improved. In addition, an adaptive threshold is constructed in the clustering analysis of the Gaussian mixture unit to integrate the discrete particle samples into similar component units, thereby reducing the complexity of the clustering calculation, and improving the real-time performance and calculation efficiency of the system's signal processing.

[0041] (5) In the present invention, the sampling samples of the continuous probability density function of the weighted point set to be constructed are stratified, and the proportion capacity of each sampling layer is set to ensure that the allocation of the number of sampling particles in the layer is reasonable, and l a Sampling the layer set and l b , l c The weights of particles in the layer group are optimized and combined to increase their probability mass, and the reliability factor of the measurement data collected by the unmanned ship multi-sensor positioning system is associated with the hierarchical sampling in the multi-sensor data algorithm. When data fusion is performed on the unmanned ship multi-sensor measurement data using the new threshold hierarchical particle filtering algorithm, the sensor measurement values ​​with a large fusion reliability factor are preferentially sampled, further increasing the reference value of the entire information sample, and finally improving the fusion positioning accuracy of the unmanned ship position data using the data fusion algorithm. [Brief description of the drawings]

[0042] [Figure 1] 1 is a flowchart of an unmanned ship positioning method based on multi-sensor data fusion of the present invention. [Diagram 2] FIG. 1 is a schematic diagram of the construction and pre-processing of an unmanned ship multi-sensor platform. [Diagram 3] 13 is a flowchart of a reliability determination process. [Figure 4] 13 is a flowchart of a consistency check and weight compensation. [Diagram 5] 13 is a flow chart of an augmentation process for elementary particle filtering data. [Figure 6] FIG. 13 is a posterior probability distribution diagram for each time. [Figure 7] 1 is a flowchart for constructing a Gaussian mixture model. [Figure 8] FIG. 13 is a density distribution diagram of the sampled particles at sampling time k=30. [Figure 9] 13 is a flow chart for setting a stratified sampling percentage capacity in relation to an adaptive threshold and a confidence factor. [Figure 10(a)] FIG. 13 is a test diagram for the algorithm comparison root mean square. [Figure 10(b)] FIG. 13 is a test diagram of the algorithm comparison standard difference. [Figure 11] This is a river channel test map. [Figure 12] FIG. 1 is a test diagram of an unmanned ship navigation trajectory. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0043] In order to make the purpose and technical solution of the embodiments of the present invention clearer, the technical solution of the embodiments of the present invention will be described clearly and completely below in conjunction with the drawings of the embodiments of the present invention. Obviously, the described embodiments are not all the embodiments, but only some of the embodiments of the present invention. Based on the described embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative labor belong to the protection scope of the present invention.

[0044] As shown in FIG. 1, the unmanned ship positioning method based on multi-sensor data fusion of the present invention first pre-processes the positioning data collected by the unmanned ship multi-sensor positioning system, then performs a confidence distance judgment on the unmanned ship positioning data and assigns a corresponding confidence factor, and at the same time performs fault inspection and weighted compensation on the positioning data, and performs a filtering process on the positioning data to realize data augmentation, and finally performs multi-sensor data fusion filtering output using a novel threshold hierarchical particle filtering algorithm, thereby realizing accurate positioning of the unmanned ship navigation trajectory.

[0045] Specifically, the following steps are taken: (1) Data preprocessing, (2) Data reliability assessment and allocation; (3) Data failure inspection compensation; (4) Data augmentation, (5) Data fusion.

[0046] As shown in FIG. 2, the method for pre-processing unmanned ship positioning data collected by an unmanned ship multi-sensor positioning system includes the following steps.

[0047] (A) Based on the latitude and longitude information of the unmanned ship's navigation trajectory collected by the positioning system sensor, the temporal and spatial standards will be unified.

[0048] (B) According to the river route along which the unmanned ship navigates, a communication environment that satisfies the effective connection of multiple nodes of the positioning system is constructed, so that the blind node located on the unmanned ship is within a network of four signal receiving nodes with known transmission signal strength and location coordinates, thereby calculating and collecting the location coordinates of the unmanned ship as it navigates the river route.

[0049] (C) Transform and unify the multi-sensor positioning system coordinates based on the Gauss-Krüger projection principle.

[0050] Information is supplemented by using positioning data collected by the unmanned ship's multi-sensors, effectively overcoming the problem of sensor signal loss due to environmental signal blockage and ensuring the validity of the data collected by the sensors.

[0051] As shown in FIG. 3, the method for making a reliability judgment and assigning a reliability factor based on data collected by multiple sensors includes the following steps: 1) According to the measurement data collected by the positioning system, a multi-sensor measurement model p i (x) and the confidence distance d between the sensor data i Calculate the p i (x), d i The calculation is as follows:

number

number

[0052] In the formula:

number

number

[0053] 2) The d obtained in step 1) according to the Gaussian probability model i Let p be the probability semantic metric between the sensor measurement data. r (Z i ) and set the sensor support reliability level, which determines the reliability of different sensor information. r (Z i ) is calculated as follows:

number

[0054] 3) The reliability of the location information collected by the multi-sensors is judged according to the sensor reliability distance, and the positioning data is divided into different reliability intervals based on the reliability of the information.

[0055] 4) Dynamic support factor β of measurement information collected by multiple sensors at time k i (k) and the Gaussian probability measurement model p i (k) norm equation is constructed, and the β i The calculation of (k) is as follows:

number

[0056] 5) The multi-sensor dynamic support factor β obtained in step (4) i Based on (k), the measurement error of the system w i Calculate the w i The calculation of is as follows: w i =z i -Aβ i (k) In the formula, A is the state transition matrix, and z i are the measurements of the unmanned ship multi-sensor positioning system, for i = 1, 2, ..., n.

[0057] 6) System measurement error w i The variance of

number

number

[0058] According to the magnitude of the probability semantic information metric value between the measurement data, the determined positioning data is divided into different confidence intervals, so that the reliable data approaches the high confidence region, and the set confidence level ε is determined by the sensor measurement variance average value

number

number

[0059] As shown in FIG. 4, the method for performing a weight compensation process on the non-matching fault data collected by the multi-sensors by performing a consistency check on the multi-sensor data includes the following steps: (I) The sensor measurement data obtained by the unmanned ship positioning system is arithmetically averaged to obtain the arithmetic mean value

number

number

number

number

number

number

number

[0060] By performing a consistency check on the unmanned ship position information collected by the multi-sensor positioning system and performing weighted compensation for faulty data, the fault tolerance performance of the data fusion algorithm is improved and the reliability of the particle set samples sampled by the basic particle filtering algorithm is ensured.

[0061] As shown in FIG. 5, the method for performing filtering on the pre-processed, inspected and compensated multi-sensor data based on a basic particle filtering algorithm includes the following steps: (i) Establish a basic particle filtering model of the unmanned ship multi-sensor positioning system, substitute the unmanned ship position information after consistency checking and variance weighting into the system model, and describe the state and measurement model of the sensor system as follows:

number

number

number

number

number

number

[0062] Particle propagation model for state transition of unmanned ship multi-sensor positioning system

number

[0063] As shown in FIG. 6 and FIG. 7, the method for constructing a Gaussian mixture model includes the following steps: (a) A multi-sensor dataset processed with a basic particle filtering algorithm and the Sigma point set

number

number

number

number

number

number

number

number

number

[0064] The importance of the new threshold hierarchical particle filtering algorithm: In the sampling process, the current measurement information is integrated into the particle set proposal distribution, making the proposal distribution closer to the actual posterior probability density, and improving the algorithm estimation performance. As can be seen from Figure 6, the constructed weighted Gaussian mixture distribution function meets the continuous multi-peak normal distribution at each time, which shows that the weighted point set mixture Gaussian distribution composed of similar component units is close to the posterior probability density function, and the function peak distribution is concentrated, which shows that the particle sample effectively represents the probability distribution characteristics after resampling.

[0065] As shown in FIG. 8 and FIG. 9, the specific steps of the method of setting a hierarchical sampling ratio capacity associated with an adaptive threshold and a reliability factor, and performing fusion filtering on the unmanned ship navigation trajectory positioning data, and outputting the unmanned ship navigation trajectory positioning information are as follows: a) After selecting the importance sampling process, the particle X with the largest weight in the discrete particle sample set is c is the clustering center, and the martensite distance D between other particles i and i Calculate the above D i is shown below,

number

number

number

number

number

[0066] h) Based on the hierarchy theory, successive probability density functions, each of which has a probability density function p(x),

number

[0067] i) each a , l b , l c Let the particle number fractional volumes of the layers be N / 4, N / 3, and N / 3.

[0068] j) Substitute the reliability factor into the weight optimization combination and perform calculation.

[0069] k)l b , l c The weights in the layer are the average

number

number

number

[0070] l) Obtaining the multi-sensor data fusion sampling result obtained in step (k).

[0071] m) outputting the data fusion sampling results obtained in step (l) from the blind node installed on the unmanned vessel in the form of a log file;

[0072] n) The PC-side coordinator node placed on the riverbed networks with the unmanned ship blind node to obtain the unmanned ship position information output by the blind node in step (u) in real time, thereby realizing positioning of the unmanned ship's navigation trajectory.

[0073] The sampling samples of the continuous probability density function of the constructed weighted point set are stratified, and the proportion capacity of each sampling layer is set to ensure that the allocation of the number of sampling particles in the layer is reasonable; and a Sampling the layer set and l b , l c The weight of the particles in the layer group is optimized and combined to increase its probability mass, and the reliability factor of the measurement data collected by the unmanned ship multi-sensor positioning system is associated with the hierarchical sampling in the multi-sensor data algorithm, so that when the new threshold hierarchical particle filtering algorithm is used to perform data fusion on the unmanned ship multi-sensor measurement data, the sensor measurement values ​​with a large reliability factor are preferentially sampled and fused, thereby increasing the reference value of the entire information sample. As can be seen from Figure 8, at time k=30, the sampling points are concentrated at the maximum value of the continuous probability density function, and the reliability factor of the sensor data is substituted into the hierarchical sampling calculation to ensure that the sampling points are concentrated at the larger value of the probability density function when the continuous probability density function is resampled, thereby increasing the reference value of the entire information sample, and further ensuring the accuracy of the measurement data fusion positioning by the new threshold hierarchical particle filtering algorithm.

[0074] In order to verify the optimization performance of the novel threshold hierarchical particle filtering algorithm of the present invention, this paper performs 30 independent tests on the root mean square error (RMSE) and standard difference (Std) on a computer simulation test system model, and compares the test results with the Extended Kalman Filter (EKF). [1] , Unscented Kalman Filter (UKF) [2],Unscented Particle Filter (UPF) [3] And Basic Particle Filtering (BPF) [4] , and thus verify the effectiveness of the improvement steps of the filtering algorithm.,The simulation time step size of the five algorithms is,k,=50, and the time interval,Δt,=1,, and the parameters of the compared algorithms are,taken from the corresponding references.

[0075] Computer simulation test model attached:

number

[0076] As can be seen from Table 1, Figure 10(a) and Figure 10(b), first, whether it is the RMSE average value or the maximum value, in the specification, TLPF is smaller than the other four algorithms, which indicates that the filtering accuracy of the novel threshold hierarchical particle filtering algorithm in this patent is the highest. This is mainly because the specification clusters discrete particles by constructing an adaptive threshold in Gaussian mixture, and performs weight optimization combination for particles in the inferior layer to improve the diversity of particle samples. From the Std average value and maximum value, it is shown that the value of TLPF is also the smallest compared to the other four algorithms, and the filtering stability of TLPF is also the best. This is mainly due to the selection of the importance function in the importance sampling process, which merges the latest measurement data into the importance function through unscented transformation to improve the reliability of the particle sample, and performs re-hierarchical sampling calculation for particles in the inferior layer in the resampling process, so that more particle samples are substituted into the sampling process, thereby ensuring the diversity of the particle sample set.

[0077] In order to verify the optimization performance of the multi-sensor data fusion algorithm of the present invention, a comparative positioning test of the multi-sensor system algorithm based on the unmanned ship platform is carried out, and 100 sets of positioning data collected by the multi-sensor system of the test river channel are selected, and the root mean square error (RMSE) and standard difference (Std) are used as the test result performance indicators, and the test results are compared with those of the extended Kalman filtering, the unscented Kalman filtering, the unscented particle filtering and the basic particle filtering. Figure 4 shows the results of filtering fusion of the unmanned ship navigation positioning trajectory data by each algorithm.

[0078] As can be seen from Table 2, Figure 11 and Figure 12, when the unmanned ship navigates the test river, due to the thick trees and narrow and narrow river channel, it has a certain environmental perturbation effect on the sensor signal transmission. The filtered positioning results of each data fusion algorithm will seriously deviate from the actual route navigated by the unmanned ship in some places. Moreover, the positioning results are prone to large jumps at the curves of the route, and will gradually deviate from the actual trajectory navigated by the unmanned ship, reducing the reliability of the positioning results. As time goes by, the error of the positioning results will become larger and larger. However, the positioning accuracy of the multi-sensor data fusion algorithm of the present invention is higher than the fusion positioning accuracy of any of the algorithms, which is mainly because the present invention first performs a fault check on the unmanned ship multi-sensor measurement data through a consistency check, and then performs a weighted compensation process on the inconsistent data to improve its reliability; when the basic particle filtering performs data augmentation on the data samples, it has more and more reliable sampling samples, which improves the processing accuracy and fault tolerance performance of the data fusion algorithm; then, a reliability check is performed on the data collected by the unmanned ship multi-sensor positioning system through a reliability check, and a corresponding positioning data reliability factor is assigned according to the magnitude of the reliability distance, which is then associated with the subsequent hierarchical sampling operation step of the sensor data fusion algorithm; finally, a novel threshold hierarchical particle filtering algorithm is used to construct a continuous probability density function of a Gaussian mixture and set an adaptive threshold to improve the clustering integration efficiency, which improves the real-time performance of the data fusion algorithm for processing the positioning data, and the reliability factor is associated with the hierarchical sampling weight calculation, so that the particle samples with high reliability factors are preferentially sampled, thereby improving the positioning accuracy of the data fusion algorithm. Compared with the other four data fusion algorithms, the multi-sensor data fusion algorithm of the present invention reduces the average positioning error by 47%, and sufficiently ensures the positioning accuracy for the unmanned ship navigation trajectory.

[0079] [Table 1]

[0080] [Table 2]

[0081] The above examples are only used to explain the technical solutions of the present invention and are not limited thereto. The present invention has been described in detail with reference to the above examples. However, those skilled in the art should understand that the specific embodiments of the present invention may be modified or equivalently substituted, and should be included within the scope of protection of the claims of the present invention without departing from the spirit and scope of the present invention.

Claims

1. 1. An unmanned ship positioning method based on multi-sensor data fusion, comprising: Step (1) of pre-processing the unmanned ship positioning data collected by the unmanned ship multi-sensor positioning system, the specific content and method steps of which are as follows: (A) Based on the latitude and longitude information of the unmanned vessel navigation trajectory collected by the positioning system sensor, unify the time and space standards for it, (B) According to the river course along which the unmanned vessel navigates, a communication environment that satisfies the effective connection of multiple nodes of the positioning system is established, so that the blind node located on the unmanned vessel is within a network consisting of four signal receiving nodes with known transmission signal strength and position coordinates, thereby calculating and collecting the position coordinates of the unmanned vessel when it navigates along the river course; (C) transforming and unifying multi-sensor positioning system coordinates based on the Gauss-Kruger projection principle; (2) performing a reliability determination and a reliability factor assignment based on data collected by multiple sensors; (3) performing a weight compensation process on the mismatched fault data collected by the positioning system by performing a consistency check on the multi-sensor data; (4) respectively filtering the pre-processed, inspected and compensated multi-sensor data based on a basic particle filtering algorithm; and (5) designing a new threshold hierarchical particle filtering algorithm by constructing a Gaussian mixture model and setting a hierarchical sampling proportion capacity associated with an adaptive threshold and a confidence factor, and further performing fusion filtering on the unmanned ship navigation trajectory positioning data, and outputting unmanned ship navigation trajectory positioning information.

2. The specific content and method steps of the step (2) of performing reliability judgment and reliability factor assignment based on data collected by multiple sensors are as follows: 1) A multi-sensor measurement model p is generated based on the measurement data collected by the positioning system. ｉ (x) and the confidence distance d ｉ Calculate the p ｉ (x), d ｉ are calculated using the following formulas, [Number 85] In the formula, x ｉ is the measurement value of the i-th sensor, μ is the true value of the measurement feature, and θ ｉ is the measurement accuracy of the i-th sensor information, and σ ｉ is the i-th sensor information measurement error, [Number 86] Each of i is a measurement value collected by the multi-sensor at time i, [Number 87] is the measurement variance mean, Z is a random variable that follows a standard normal distribution, i = 1, 2, ..., n, 2) The d obtained in step 1) according to the Gaussian probability model ｉ Let p be the probability semantic metric between sensor measurement data. ｒ (Z ｉ ) and set a sensor support reliability level, thereby judging the reliability of different sensor information, ｒ (Z ｉ ) is calculated using the following formula: [Number 88] In the formula, Z ｉ is the multi-sensor measurement data, ε is the confidence level, and K is the coefficient of variation of the sampling probability interval; 3) determining the reliability of the location information collected by the multi-sensor according to the sensor reliability distance, and dividing the positioning data into different reliability intervals according to the reliability of the information; 4) Dynamic support factor β of measurement information collected by multiple sensors at time k ｉ (k) and the Gaussian probability measurement model p ｉ (k) and construct a norm equation for the β ｉ (k) is calculated using the following formula: [Number 89] Frobenius norm, k = 1, 2, ..., T, i = 1, 2, ..., n, 5) The multi-sensor dynamic support factor β obtained in step 4) ｉ Based on (k), the measurement error of the system w ｉ Calculate the w ｉ is calculated using the following formula: w ｉ =z ｉ -Aβ ｉ (k) In the formula, A is the state transition matrix, and z ｉ are the measurements of the unmanned ship multi-sensor positioning system, i = 1, 2, ..., n; 6) System measurement error w ｉ The variance of [Number 90] Confidence factor for measurement data using [Number 91] 2. The unmanned ship positioning method based on multi-sensor data fusion according to claim 1, characterized in that i is the reliability factor of the sensor measurement data, and i=1, 2, ..., n.

3. The specific content and method steps of step (3) of performing a consistency check on the multi-sensor data to perform weight compensation processing on the mismatched fault data collected by the positioning system are as follows: (I) The sensor measurement data obtained by the unmanned ship positioning system is arithmetically averaged to obtain the arithmetic mean value [Number 92] In the formula, x ｉ is the sensor measurement information, i = 1, 2, ..., n, (II) The arithmetic mean value of the sensor measurement data obtained in step (I) and the subsequent sampling value x of the positioning system ｈ Subtract and (III) setting a system required error, and when the difference value between the arithmetic mean value of the sensor measurement data and the subsequent sampling value of the positioning system is smaller than the system required error, determining that the unmanned ship position data collected by the multi-sensor positioning system has consistency and is reliable data; when the difference value is larger than the system required error, performing variance weighting compensation on the sampling data, thereby meeting the sampling demand of data samples through basic particle filtering; (IV) Arithmetic mean value of multi-sensor measurement data [Number 93] When an unmanned vessel multi-sensor positioning system measures the unmanned vessel navigation trajectory at different positions in the same space, the information measurement variance of the i-th sensor system is [Number 94] (V) According to the data set recorded by the unmanned ship multi-sensor positioning system performing m measurements on the unmanned ship, the j-th measurement data of the i-th sensor is calculated as x ｉｊ Let x ｉｊ In step (IV), the sensor measurement information x ｉ This replaces the dataset information variance obtained by multiple measurements. [Number 95] In the formula, i=1, 2, ..., n and j=1, 2, ..., m. (VI) Based on the location information data set distribution calculated in step (V), a fusion weight k of the mismatched fault data in step (III) is calculated. ｉ Define the k ｉ is calculated using the following formula: [Number 96] (VII) The calculated fusion weight k ｉ Based on the discrepancy data [Number 97] The sensor measurement data is weighted and compensated to meet the basic particle filtering requirements. [Number 98] 2. The method for positioning an unmanned ship based on multi-sensor data fusion according to claim 1, wherein: [Number 99]

4. The specific content and method steps of step (4) of performing filtering processes on the multi-sensor data after pre-processing, inspection and compensation according to a basic particle filtering algorithm are as follows: (i) Establish a basic particle filtering model of the unmanned ship multi-sensor positioning system, substitute the unmanned ship position information after consistency checking and variance weighting into the system model, and describe the state and measurement model of the sensor system as follows: [Number 100] In the formula, x ｋ is the position prediction value of the sensor system at time k, and x ｈ is the subsequent sampled value of the sensor, [Number 101] is the variance-weighted sensor measurement, and z ｋ is the drone position measurement at time k, and λ ｋ is the estimated noise, and v ｋ is the measurement noise, (ii) Initialize the particle set sample to obtain the prior density p(x ０ ) to randomly sample the initial particle set [00102] (iii) randomly draw a sample of N particles from the importance density function; (iv) Weights of the sampling particles [Number 103] are the weights of the sampling particles, i = 1, 2, ..., N, (v) normalizing the importance weights; (vi) The effective number of particles in the basic particle filtering algorithm, N ｅｆｆ and calculate a threshold N ｔｈ Compared with N ｅｆｆ <N ｔｈ Resampling is performed in the case of N ｅｆｆ is calculated using the following formula: [Equation 104] (vii) outputting a state and estimating a part of the measurement information collected by the unmanned ship multi-sensor positioning system filtered; [Number 105] 2. The method for positioning an unmanned ship based on multi-sensor data fusion according to claim 1, wherein: [Number 106]

5. The specific content and method steps of the step (5) of constructing a Gaussian mixture model are as follows: (a) Extracting a multi-sensor data set that has been processed with a basic particle filtering algorithm and extracting the Sigma point set [Equation 107] is the Sigma point set after the unscented transformation, and n ａ is the dimension of the Sigma point, λ is the scale parameter, (b) Fusing the latest measurement information with the resulting set of Sigma sampling points and determining the system state [Number 108] In the formula, k ｋ is the Kalman gain, and z ｋ is the measurement information, [Number 109] is the covariance measured by the weighted Sigma point set, (c) the multi-sensor positioning system state obtained in step (b); [Equation 110] and sampling from the proposed distribution constructed, (d) Posterior probability density function for time step size k according to the Gaussian mixture model [Equation 111] is the i-th component in the Gaussian mixture model, C(k) is the number of component units in the discrete samples, and ξ is the discrete point component weight. (e) The discrete sampling points sampled in step (c) and their corresponding weights. [Equation 112] , and the Gaussian mixture component unit of step (d) is fused to construct a continuous probability density function [Equation 113] where p(k) is the covariance of the discrete particle filtering distribution, [Equation 114] is the mean of the discrete particle filtering distribution, h is a standardization constant, and n ｘ is the particle distribution dimension, (f) employing clustering analysis to obtain the continuous posterior probability density function of step (e); [Number 115] The unmanned ship positioning method based on multi-sensor data fusion according to claim 1, characterized in that: the Gaussian mixture similarity units in are jointly processed.

6. The specific content and method steps of the step (5) of setting the adaptive threshold are as follows: (g) After selecting the importance sampling process, the particle X with the largest weight in the discrete particle sample set is ｃ is the clustering center, and the martensite distance D between other particles i and ｉ Calculate the above D ｉ is shown below, [Equation 116] is the probability density of particle i, S is the covariance matrix, (h) The number of effective particle samples in the clustering unit, N ｅ Calculate the above N ｅ is calculated using the following formula: [Equation 117] is the particle probability density covariance, (i) construct a threshold T, where T is [Number 118] In the formula, T ０ is the initial threshold value, and k ｅ is the proportion coefficient, R is the number of classifications, (j) The number of effective particle samples N obtained in step (h) ｅ Substituting into the threshold T, the adaptive threshold T ｃ and constructing said T ｃ is shown below, [Equation 119] (k) D ｉ and the adaptive threshold T ｃ Compare with D ｉ T ｃ If the particle is smaller than ｉ T ｃ If ,is larger than ,, then skip this particle and cluster other particles, (m) selecting the particle with the largest weight from the other particle samples as a clustering center, and repeating step (k) until clustering is completed; (n) A continuous probability density function of the particle set constructed according to the component units after clustering. [Number 120] In the formula, β ｉ is the probability mass of similar component unit component i, and γ ｉ is the average value of component unit component i, and p ｉ 2. The method for unmanned ship positioning based on multi-sensor data fusion according to claim 1, wherein: i is the covariance of component unit component i, where i=1, 2, ..., n.

7. The specific content and method steps of the step (5) of relating the reliability factor to the hierarchical sampling proportion capacity, and further performing fusion filtering on the unmanned ship navigation trajectory positioning data, and outputting the unmanned ship navigation trajectory positioning information are as follows: (o) Based on the hierarchy theory, a series of probability density functions, each of which is p(x), [Number 121] Divide into l layers, and divide the group layers into one weighted dominant layer and two subordinate layers according to the size of the probability mass, and ａ , l ｂ , l ｃ Define it as: (p) each l ａ , l ｂ , l ｃ The particle number fractional volumes of the layers are N / 4, N / 3, and N / 3, (q) calculating the reliability factor by substituting it into the weight optimization combination; (r)l ｂ , l ｃ The weights in the layer are the average [Number 122] The weights of the particles after the optimization are [Number 123] and performing stratified sampling on the sample data; [Number 124] (s) acquiring the multi-sensor data fusion sampling result obtained in step (r); (t) fusing the data acquired in step (s) into a sampling result and outputting it in the form of a log file from the blind node installed on the unmanned vessel; The unmanned ship positioning method based on multi-sensor data fusion as described in claim 1, characterized in that (u) the PC side coordinator node placed on the riverbed networks with the unmanned ship blind node to obtain the unmanned ship position information output by the blind node in real time in step (u), thereby realizing positioning of the unmanned ship navigation trajectory.

Images