[go: up one dir, main page]

CN115790629A - A detection method for the path tracking accuracy of an autonomous sightseeing vehicle - Google Patents

A detection method for the path tracking accuracy of an autonomous sightseeing vehicle Download PDF

Info

Publication number
CN115790629A
CN115790629A CN202211458202.1A CN202211458202A CN115790629A CN 115790629 A CN115790629 A CN 115790629A CN 202211458202 A CN202211458202 A CN 202211458202A CN 115790629 A CN115790629 A CN 115790629A
Authority
CN
China
Prior art keywords
sightseeing vehicle
automatic driving
state vector
self
driving sightseeing
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Granted
Application number
CN202211458202.1A
Other languages
Chinese (zh)
Other versions
CN115790629B (en
Inventor
张伟
赵奉奎
张永举
张健
黄伟华
董洲
吴天余
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Special Equipment Safety Supervision Inspection Institute of Jiangsu Province
Original Assignee
Special Equipment Safety Supervision Inspection Institute of Jiangsu Province
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Special Equipment Safety Supervision Inspection Institute of Jiangsu Province filed Critical Special Equipment Safety Supervision Inspection Institute of Jiangsu Province
Priority to CN202211458202.1A priority Critical patent/CN115790629B/en
Publication of CN115790629A publication Critical patent/CN115790629A/en
Application granted granted Critical
Publication of CN115790629B publication Critical patent/CN115790629B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

Landscapes

  • Optical Radar Systems And Details Thereof (AREA)
  • Navigation (AREA)

Abstract

The invention discloses a method for detecting the path tracking precision of an automatic driving sightseeing vehicle, which obtains the path tracking precision of the automatic driving sightseeing vehicle by sequentially obtaining the tracking error of the automatic driving sightseeing vehicle to each target point and according to the maximum value of the tracking errors of all the target points. Meanwhile, when tracking errors of all target points are obtained, the adaptive setting of Kalman filtering algorithm coefficients is realized through the setting of an adaptive coefficient switch function, further the dynamic adaptive adjustment of predicted values and observation value weights in the process of Kalman filtering algorithm state vector optimal estimation is realized, and the problem of Kalman gain divergence in the recursion process is solved.

Description

一种自动驾驶观光车路径跟踪精度检测方法A detection method for the path tracking accuracy of an autonomous sightseeing vehicle

技术领域technical field

本发明涉及一种路径跟踪精度的检测方法，尤其涉及一种自动驾驶观光车路径跟踪精度检测方法。The invention relates to a detection method of path tracking accuracy, in particular to a detection method of path tracking accuracy of an automatic driving sightseeing vehicle.

背景技术Background technique

自动驾驶观光车在提升主动安全性、提高交通效率以及降低能源消耗方面具有巨大的优势，因此自动驾驶技术成为当前行业研究热点。路径跟踪是自动驾驶观光车的关键功能之一，实现路径精确跟踪，是观光车自动驾驶的基础，可保障行车安全，提高车辆稳定性。因此，路径跟踪精度是衡量自动驾驶观光车性能的一个关键指标。为了实现自动驾驶观光车路径跟踪精度的检测，需要对自动驾驶观光车的位置实时估算，用以计算自动驾驶观光车实际行驶路径与目标路径的横向或纵向误差。Self-driving sightseeing cars have huge advantages in improving active safety, improving traffic efficiency and reducing energy consumption, so self-driving technology has become a current research hotspot in the industry. Path tracking is one of the key functions of self-driving sightseeing cars. Accurate path tracking is the basis for automatic driving of sightseeing cars, which can ensure driving safety and improve vehicle stability. Therefore, path tracking accuracy is a key indicator to measure the performance of autonomous sightseeing vehicles. In order to realize the detection of the path tracking accuracy of the self-driving sightseeing car, it is necessary to estimate the position of the self-driving sightseeing car in real time to calculate the lateral or vertical error between the actual driving path of the self-driving sightseeing car and the target path.

当前观光车定位主要采用全球定位系统GPS实现。然而GPS存在一定的误差，更新频率仅有约10Hz，观光车在有信号遮挡的环境中容易出现GPS信号丢失的问题，因此仅使用GPS无法满足自动驾驶观光车定位的需求。At present, the positioning of sightseeing cars is mainly realized by the global positioning system GPS. However, there is a certain error in GPS, and the update frequency is only about 10Hz. The sight-seeing car is prone to GPS signal loss in an environment with signal occlusion. Therefore, only using GPS cannot meet the needs of self-driving sightseeing car positioning.

惯性导航系统INS更新频率可达100Hz，且不易受外界干扰，然而其定位原理决定了INS存在误差累积的问题。The update frequency of the inertial navigation system INS can reach 100Hz, and it is not susceptible to external interference. However, its positioning principle determines that INS has the problem of error accumulation.

当前主流的定位方法是采用卡尔曼滤波算法融合GPS和INS，以INS的输出预测自动驾驶观光车当前位置，利用GPS输出的观测信息对所预测的位置进行更新，实现对自动驾驶观光车位置的最优估计。卡尔曼滤波算法对状态向量的最优估计是对预测值和观测值的加权处理，随着递推过程的继续，对卡尔曼增益，即权值的计算会变的越来越不准确，因此需要对卡尔曼滤波算法进行改进。The current mainstream positioning method is to use the Kalman filter algorithm to fuse GPS and INS, use the output of INS to predict the current position of the self-driving sightseeing car, and use the observation information output by GPS to update the predicted position to realize the location of the self-driving sightseeing car. best estimate. The optimal estimation of the state vector by the Kalman filter algorithm is the weighted processing of the predicted value and the observed value. As the recursive process continues, the calculation of the Kalman gain, that is, the weight value, will become more and more inaccurate. Therefore, The Kalman filter algorithm needs to be improved.

为此，提供一种新的技术方案改善上述问题，以提高自动驾驶观光车路径跟踪检验精度，是本领域技术人员急需解决的问题。Therefore, it is an urgent problem for those skilled in the art to provide a new technical solution to improve the above-mentioned problems so as to improve the accuracy of the path tracking inspection of the self-driving sightseeing car.

发明内容Contents of the invention

针对现有技术中存在的问题，本发明提供一种自动驾驶观光车路径跟踪精度检测方法，自适应调整卡尔曼滤波算法中的相关系数，提高自动驾驶观光车定位精度，进而提高自动驾驶观光车路径跟踪检验精度。Aiming at the problems existing in the prior art, the present invention provides a method for detecting the path tracking accuracy of the self-driving sightseeing car, which adaptively adjusts the correlation coefficient in the Kalman filter algorithm, improves the positioning accuracy of the self-driving sightseeing car, and further improves the accuracy of the self-driving sightseeing car. Path-following test accuracy.

本发明的自动驾驶观光车路径跟踪精度检测方法，包括自动驾驶观光车路径跟踪精度检测方法，包括以下步骤；The method for detecting the path tracking accuracy of the self-driving sightseeing car of the present invention includes the method for detecting the path tracking accuracy of the self-driving sightseeing car, comprising the following steps;

在规划的路径上采集若干目标点的坐标数据；Collect the coordinate data of several target points on the planned path;

在自动驾驶观光车按所述路径自动行驶过程中，依次获取自动驾驶观光车对各目标点的跟踪误差；During the automatic driving process of the self-driving sightseeing car according to the described path, the tracking errors of the self-driving sightseeing car to each target point are sequentially obtained;

多个跟踪误差中，误差值最大的即为自动驾驶光观车的路径跟踪精度；Among multiple tracking errors, the one with the largest error value is the path tracking accuracy of the self-driving sightseeing car;

其特征在于：自动驾驶观光车对某目标点的跟踪误差的获取方法，包括以下步骤；It is characterized in that: the method for obtaining the tracking error of the automatic driving sightseeing car to a certain target point includes the following steps;

步骤S1；Step S1;

在自动驾驶观光车接近及远离目标点的过程中，每间隔预定时间获取自动驾驶观光车当前位置的最优估计值；In the process of the self-driving sightseeing car approaching and moving away from the target point, the optimal estimate value of the current position of the self-driving sightseeing car is obtained at predetermined intervals;

步骤S2；Step S2;

根据步骤S1中获取的自动驾驶观光车当前位置的最优估计值，计算其与目标点之间的距离；Calculate the distance between it and the target point according to the optimal estimated value of the current position of the self-driving sightseeing car obtained in step S1;

步骤S3；Step S3;

由步骤S1及步骤S2获取的多个最优估计值与目标点之间的距离中，距离最小的即为自动驾驶观光车对该目标点的跟踪误差；Among the distances between multiple optimal estimated values and the target point obtained by step S1 and step S2, the smallest distance is the tracking error of the target point by the self-driving sightseeing car;

步骤S1中，某k时刻，自动驾驶光观车当前位置最优估计值的获取方法，包括如下步骤；In step S1, at a certain k moment, the method for obtaining the optimal estimated value of the current position of the self-driving sightseeing car includes the following steps;

步骤S11；Step S11;

根据自动驾驶观光车的状态参数建立状态向量

Figure 640513DEST_PATH_IMAGE001
和观测向量
Figure 502421DEST_PATH_IMAGE002
，状态向量包括自 动驾驶观光车的位置、速度和航向角参数，状态向量
Figure 998256DEST_PATH_IMAGE001
和观测向量
Figure 575999DEST_PATH_IMAGE002
的表达式如下； Establish the state vector according to the state parameters of the self-driving sightseeing car
Figure 640513DEST_PATH_IMAGE001
and observation vector
Figure 502421DEST_PATH_IMAGE002
, the state vector includes the position, speed and heading angle parameters of the self-driving sightseeing car, and the state vector
Figure 998256DEST_PATH_IMAGE001
and observation vector
Figure 575999DEST_PATH_IMAGE002
The expression of is as follows;

Figure 954022DEST_PATH_IMAGE003
+
Figure 540992DEST_PATH_IMAGE004
Figure 954022DEST_PATH_IMAGE003
+
Figure 540992DEST_PATH_IMAGE004

Figure 15967DEST_PATH_IMAGE005
Figure 15967DEST_PATH_IMAGE005

预测k时刻自动驾驶观光车先验状态向量

Figure 702294DEST_PATH_IMAGE006
和先验协估计方差
Figure 879198DEST_PATH_IMAGE007
，预测公 式如下； Predict the prior state vector of the self-driving sightseeing car at time k
Figure 702294DEST_PATH_IMAGE006
and prior co-estimating the variance
Figure 879198DEST_PATH_IMAGE007
, the prediction formula is as follows;

Figure 738700DEST_PATH_IMAGE008
Figure 738700DEST_PATH_IMAGE008

Figure 599340DEST_PATH_IMAGE009
Figure 599340DEST_PATH_IMAGE009

其中，

Figure 710429DEST_PATH_IMAGE010
为从第k-1时刻到第k时刻自动驾驶观光车的状态转移矩阵，
Figure 843470DEST_PATH_IMAGE011
为k-1时刻自动驾驶观光车的后验状态向量，
Figure 444347DEST_PATH_IMAGE012
为k时刻自动驾驶光观车的控制矩阵，
Figure 425073DEST_PATH_IMAGE013
为k 时刻自动驾驶光观车的控制输入向量，T(s)为自适应系数开关函数，
Figure 968049DEST_PATH_IMAGE014
为k-1时刻自 动驾驶观光车的后验估计协方差，
Figure 604698DEST_PATH_IMAGE015
为
Figure 71583DEST_PATH_IMAGE010
的转置矩阵，
Figure 828186DEST_PATH_IMAGE016
为k时刻，服从均值为0正态 分布 in,
Figure 710429DEST_PATH_IMAGE010
is the state transition matrix of the self-driving sightseeing car from the k -1 moment to the k -th moment,
Figure 843470DEST_PATH_IMAGE011
is the posterior state vector of the self-driving sightseeing car at time k -1,
Figure 444347DEST_PATH_IMAGE012
is the control matrix of the self-driving sightseeing car at time k ,
Figure 425073DEST_PATH_IMAGE013
is the control input vector of the self-driving sightseeing car at time k , T( s ) is the adaptive coefficient switching function,
Figure 968049DEST_PATH_IMAGE014
is the posterior estimated covariance of the self-driving sightseeing car at time k -1,
Figure 604698DEST_PATH_IMAGE015
for
Figure 71583DEST_PATH_IMAGE010
The transpose matrix of
Figure 828186DEST_PATH_IMAGE016
is time k , obeying the normal distribution with mean value 0

的系统噪声向量，其中，T(s)为自适应系数开关函数，如下所示；The system noise vector of , where T (s) is the adaptive coefficient switching function, as shown below;

T(s)=

Figure 89534DEST_PATH_IMAGE017
T (s) =
Figure 89534DEST_PATH_IMAGE017

式中s为自适应系数，超参数s 0为大于1的自适应系数阈值；In the formula , s is the adaptive coefficient, and the hyperparameter s 0 is the adaptive coefficient threshold greater than 1;

步骤S12；Step S12;

根据观测值更新步骤S11中获取的先验状态向量

Figure 479059DEST_PATH_IMAGE018
及先验斜估计方差
Figure 952896DEST_PATH_IMAGE007
， 更新公式如下所示； Update the prior state vector obtained in step S11 according to the observed value
Figure 479059DEST_PATH_IMAGE018
and a prior skew estimated variance
Figure 952896DEST_PATH_IMAGE007
, the update formula is as follows;

Figure 95165DEST_PATH_IMAGE019
Figure 95165DEST_PATH_IMAGE019

Figure 792993DEST_PATH_IMAGE020
Figure 792993DEST_PATH_IMAGE020

其中，

Figure 607497DEST_PATH_IMAGE021
为k时刻自动驾驶观光车的后验状态向量，
Figure 399872DEST_PATH_IMAGE022
为卡尔曼增益，
Figure 944117DEST_PATH_IMAGE002
为观测 值，
Figure 281689DEST_PATH_IMAGE023
为k时刻状态变量到观测变量的转换矩阵，
Figure 645805DEST_PATH_IMAGE024
为k时刻后验估计协方差； in,
Figure 607497DEST_PATH_IMAGE021
is the posterior state vector of the self-driving sightseeing car at time k ,
Figure 399872DEST_PATH_IMAGE022
for the Kalman gain,
Figure 944117DEST_PATH_IMAGE002
is the observed value,
Figure 281689DEST_PATH_IMAGE023
is the transformation matrix from the state variable to the observed variable at time k ,
Figure 645805DEST_PATH_IMAGE024
Posteriori estimates covariance for time k ;

其中，

Figure 179555DEST_PATH_IMAGE025
-1 in,
Figure 179555DEST_PATH_IMAGE025
-1

式中，

Figure 515989DEST_PATH_IMAGE026
为服从均值为0正态分布的高斯白噪声；In the formula,
Figure 515989DEST_PATH_IMAGE026
Gaussian white noise that obeys a normal distribution with a mean of 0;

所述后验状态向量为k时刻自动驾驶观光车状态向量的当前最优估计向量，其中包含自动驾驶观光车当前位置的最优估计值。The posterior state vector is the current optimal estimation vector of the state vector of the self-driving sightseeing car at time k , which contains the optimal estimated value of the current position of the self-driving sightseeing car.

进一步的，本发明的自动驾驶观光车路径跟踪精度检测方法，步骤S11中自适应系数s通过如下公式计算；Further, in the method for detecting the path tracking accuracy of the self-driving sightseeing car of the present invention, the self-adaptive coefficient s is calculated by the following formula in step S11;

Figure 352358DEST_PATH_IMAGE027
Figure 352358DEST_PATH_IMAGE027

其中，N为超参数，表示所取的k时刻之前的时间窗宽度；Among them, N is a hyperparameter, representing the time window width before the selected k moment;

Figure 938191DEST_PATH_IMAGE028
为第k时刻观测向量z k 与预测值
Figure 569242DEST_PATH_IMAGE029
之间的偏差，计算公式如下；
Figure 938191DEST_PATH_IMAGE028
is the observed vector z k and the predicted value at the kth moment
Figure 569242DEST_PATH_IMAGE029
The deviation between is calculated as follows;

Figure 71767DEST_PATH_IMAGE030
Figure 71767DEST_PATH_IMAGE030

Figure 282300DEST_PATH_IMAGE031
为
Figure 824271DEST_PATH_IMAGE028
的转置；
Figure 282300DEST_PATH_IMAGE031
for
Figure 824271DEST_PATH_IMAGE028
the transposition of

Tr()为矩阵迹计算函数，下标j对应由0开始至N-1号自动驾驶观光车驶过的各标记点序号；Tr() is a matrix trace calculation function, and the subscript j corresponds to the serial number of each marker point from 0 to N -1 automatic driving sightseeing car passing by;

当s≥ s 0时，更新步骤S12会增大自动驾驶观光车先验状态向量的偏差，从而增大更新阶段观测信息的权重。When s ≥ s 0 , the update step S12 will increase the deviation of the prior state vector of the self-driving sightseeing car, thereby increasing the weight of the observation information in the update stage.

进一步的，本发明的自动驾驶观光车路径跟踪精度检测方法，所述目标点由RTK高精度定位设备采集。Furthermore, in the method for detecting the path tracking accuracy of the self-driving sightseeing car of the present invention, the target points are collected by RTK high-precision positioning equipment.

进一步的，本发明的自动驾驶观光车路径跟踪精度检测方法，所述自动驾驶观光车上安装有GPS/INS组合导航系统，所述观测值由GPS/INS组合导航系统获取。Further, in the method for detecting the path tracking accuracy of the self-driving sightseeing car of the present invention, the self-driving sightseeing car is equipped with a GPS/INS integrated navigation system, and the observation value is obtained by the GPS/INS integrated navigation system.

进一步的，本发明的自动驾驶观光车路径跟踪精度检测方法，所述先验状态向量及后验状态向量包括自动驾驶观光车的位置、速度和姿态参数。Furthermore, in the method for detecting the path tracking accuracy of the self-driving sightseeing car of the present invention, the prior state vector and the posterior state vector include the position, speed and attitude parameters of the self-driving sightseeing car.

该自动驾驶观光车路径跟踪精度检测方法的优点在于，其通过依次获取自动驾驶光观车对各目标点的跟踪误差，并根据所有目标点跟踪误差的最大值，得到自动驾驶光观车的路径跟踪精度。同时，在对各目标点跟踪误差的获取时，通过自适应系数开关函数的设定，实现了卡尔曼滤波算法系数的自适应设定，进而实现了卡尔曼滤波算法状态向量最优估计过程中，预测值和观测值权重的动态自适应调整，克服了递推过程中卡尔曼增益发散的问题。The advantage of the method for detecting the path tracking accuracy of the self-driving sightseeing car is that it obtains the tracking errors of the self-driving sightseeing car on each target point in sequence, and obtains the path of the self-driving sightseeing car according to the maximum value of the tracking errors of all target points Tracking accuracy. At the same time, when obtaining the tracking error of each target point, through the setting of the adaptive coefficient switch function, the adaptive setting of the coefficient of the Kalman filter algorithm is realized, and then the optimal estimation process of the state vector of the Kalman filter algorithm is realized. , the dynamic adaptive adjustment of the predicted value and the weight of the observed value overcomes the problem of Kalman gain divergence in the recursive process.

上述说明仅是本发明技术方案的概述，为了能够更清楚地了解本发明的技术手段，并依照说明书的内容予以具体实施，以下以本发明的实施例对其进行详细说明。The above description is only an overview of the technical solutions of the present invention. In order to understand the technical means of the present invention more clearly and implement them in accordance with the contents of the description, the embodiments of the present invention will be described in detail below.

附图说明Description of drawings

图1是本发明自动驾驶观光车路径跟踪精度检测方法的流程图。Fig. 1 is a flow chart of the method for detecting the path tracking accuracy of the self-driving sightseeing car of the present invention.

图2是自动驾驶观光车需要跟踪的目标路径及目标点的示意图。Figure 2 is a schematic diagram of the target path and target points that the self-driving sightseeing car needs to track.

图中，路径1；路径中心线2；目标点3。In the figure, path 1; path centerline 2; target point 3.

具体实施方式Detailed ways

下面结合附图和实施例，对本发明的具体实施方式作进一步详细描述。以下实施例用于说明本发明，但不用来限制本发明的范围。The specific implementation manners of the present invention will be further described in detail below in conjunction with the accompanying drawings and embodiments. The following examples are used to illustrate the present invention, but are not intended to limit the scope of the present invention.

参见图1至2，本实施例的自动驾驶观光车路径跟踪精度检测方法，包括以下步骤；Referring to Figures 1 to 2, the method for detecting the path tracking accuracy of the self-driving sightseeing car of the present embodiment includes the following steps;

规划自动驾驶观光车要跟踪的路径1；Plan the path 1 to be tracked by the self-driving sightseeing car;

利用RTK高精度定位设备采集路径中心线2上的M个目标点3的坐标信息，M个目标点分别编号为1号，2号，…，M号，起点处为1号目标点，终点处为M号目标点，将M个目标点的坐标信息输入自动驾驶观光车的自动驾驶控制系统中；Use RTK high-precision positioning equipment to collect the coordinate information of M target points 3 on the path centerline 2. The M target points are respectively numbered No. 1, No. 2, ..., M. The starting point is the No. 1 target point, and the end point is the No. 1 target point. For the M target point, input the coordinate information of the M target points into the automatic driving control system of the self-driving sightseeing car;

在车上安装GPS/INS组合导航系统，GPS/INS组合导航系统采用自适应卡尔曼滤波算法实现高精度定位；Install GPS/INS integrated navigation system on the vehicle, GPS/INS integrated navigation system uses adaptive Kalman filter algorithm to achieve high-precision positioning;

将自动驾驶观光车驾驶至1号目标点，自适应卡尔曼滤波算法利用GPS/INS组合导航系统输出的数据对自动驾驶观光车当前位置做最优估计，系统根据自动驾驶观光车的位置、速度和航向角等建立状态向量，系统状态空间表达式如下所示；Drive the self-driving sightseeing car to the No. 1 target point, and the adaptive Kalman filter algorithm uses the data output by the GPS/INS integrated navigation system to make an optimal estimate of the current position of the self-driving sightseeing car. Establish a state vector with the heading angle, etc., and the system state space expression is as follows;

Figure 247293DEST_PATH_IMAGE003
+
Figure 338746DEST_PATH_IMAGE004
Figure 247293DEST_PATH_IMAGE003
+
Figure 338746DEST_PATH_IMAGE004

Figure 657863DEST_PATH_IMAGE032
Figure 657863DEST_PATH_IMAGE032

式中，

Figure 952709DEST_PATH_IMAGE001
，
Figure 976160DEST_PATH_IMAGE033
分别为第k时刻和第k-1时刻自动驾驶观光车的状态向量，
Figure 390960DEST_PATH_IMAGE034
为 从第k-1时刻到第k时刻的状态转移矩阵，
Figure 880979DEST_PATH_IMAGE035
为控制矩阵，
Figure 194279DEST_PATH_IMAGE036
为k时刻控制输入向量，
Figure 677213DEST_PATH_IMAGE004
为k-1时刻系统噪声向量，噪声服从均值为0的正态分布，协方差矩阵表示为
Figure 290729DEST_PATH_IMAGE037
，
Figure 686069DEST_PATH_IMAGE002
为第k时 刻的观测向量，
Figure 470354DEST_PATH_IMAGE023
为第k时刻状态变量到观测变量的转换矩阵，
Figure 38870DEST_PATH_IMAGE038
为第k时刻的观测噪声向 量，为高斯白噪声，服从均值为0的正态分布，协方差矩阵表示为
Figure 913416DEST_PATH_IMAGE026
。 In the formula,
Figure 952709DEST_PATH_IMAGE001
,
Figure 976160DEST_PATH_IMAGE033
are the state vectors of the self-driving sightseeing car at the kth moment and the k-1th moment, respectively,
Figure 390960DEST_PATH_IMAGE034
is the state transition matrix from the k-1th moment to the kth moment,
Figure 880979DEST_PATH_IMAGE035
is the control matrix,
Figure 194279DEST_PATH_IMAGE036
is the control input vector at time k ,
Figure 677213DEST_PATH_IMAGE004
is the system noise vector at time k-1 , the noise obeys a normal distribution with a mean value of 0, and the covariance matrix is expressed as
Figure 290729DEST_PATH_IMAGE037
,
Figure 686069DEST_PATH_IMAGE002
is the observation vector at the kth moment,
Figure 470354DEST_PATH_IMAGE023
is the transition matrix from the state variable to the observed variable at the kth moment,
Figure 38870DEST_PATH_IMAGE038
is the observation noise vector at the kth moment, which is Gaussian white noise and obeys a normal distribution with a mean value of 0, and the covariance matrix is expressed as
Figure 913416DEST_PATH_IMAGE026
.

卡尔曼滤波算法对自动驾驶观光车位置的最优估计分为预测和更新两步，在预测 步骤中，预测k时刻自动驾驶观光车先验状态向量

Figure 56821DEST_PATH_IMAGE006
和先验协估计方差
Figure 813556DEST_PATH_IMAGE007
，预测公 式如式所示， The optimal estimation of the position of the self-driving sightseeing car by the Kalman filter algorithm is divided into two steps: prediction and updating. In the prediction step, the prior state vector of the self-driving sightseeing car at time k is predicted
Figure 56821DEST_PATH_IMAGE006
and prior co-estimating the variance
Figure 813556DEST_PATH_IMAGE007
, the prediction formula is shown in the formula,

Figure 123446DEST_PATH_IMAGE008
Figure 123446DEST_PATH_IMAGE008

Figure 898504DEST_PATH_IMAGE009
Figure 898504DEST_PATH_IMAGE009

其中，

Figure 494701DEST_PATH_IMAGE010
为从第k-1时刻到第k时刻自动驾驶观光车的状态转移矩阵，
Figure 195855DEST_PATH_IMAGE011
为k-1时刻自动驾驶观光车的后验状态向量，
Figure 558703DEST_PATH_IMAGE012
为k时刻自动驾驶光观车的控制矩阵，
Figure 470158DEST_PATH_IMAGE013
为k 时刻自动驾驶光观车的控制输入向量，T(s)为自适应系数开关函数，
Figure 237257DEST_PATH_IMAGE014
为k-1时刻自 动驾驶观光车的后验估计协方差，
Figure 703005DEST_PATH_IMAGE015
为
Figure 885855DEST_PATH_IMAGE010
的转置矩阵，
Figure 838768DEST_PATH_IMAGE016
为k时刻，服从均值为0正态 分布 in,
Figure 494701DEST_PATH_IMAGE010
is the state transition matrix of the self-driving sightseeing car from the k -1 moment to the k -th moment,
Figure 195855DEST_PATH_IMAGE011
is the posterior state vector of the self-driving sightseeing car at time k -1,
Figure 558703DEST_PATH_IMAGE012
is the control matrix of the self-driving sightseeing car at time k ,
Figure 470158DEST_PATH_IMAGE013
is the control input vector of the self-driving sightseeing car at time k , T( s ) is the adaptive coefficient switching function,
Figure 237257DEST_PATH_IMAGE014
is the posterior estimated covariance of the self-driving sightseeing car at time k -1,
Figure 703005DEST_PATH_IMAGE015
for
Figure 885855DEST_PATH_IMAGE010
The transpose matrix of
Figure 838768DEST_PATH_IMAGE016
is time k , obeying the normal distribution with mean value 0

的系统噪声向量，其中，T(s)为自适应系数开关函数，如下所示；The system noise vector of , where T (s) is the adaptive coefficient switching function, as shown below;

T(s)=

Figure 714451DEST_PATH_IMAGE017
T (s) =
Figure 714451DEST_PATH_IMAGE017

式中s为自适应系数，超参数s 0为大于1的自适应系数阈值；In the formula , s is the adaptive coefficient, and the hyperparameter s 0 is the adaptive coefficient threshold greater than 1;

在更新步骤中，系统根据GPS卫星星历中获取的卫星位置、速度等信息，观测自动 驾驶观光车对各卫星伪距和多普勒频移值等，根据观测值和预测值偏差

Figure 385604DEST_PATH_IMAGE039
，更 新先验估计向量，如式所示； In the update step, the system observes the pseudo-range and Doppler frequency shift values of each satellite by the self-driving sightseeing car according to the satellite position and speed information obtained in the GPS satellite ephemeris, and according to the deviation between the observed value and the predicted value
Figure 385604DEST_PATH_IMAGE039
, to update the prior estimation vector, as shown in the formula;

Figure 434463DEST_PATH_IMAGE019
Figure 434463DEST_PATH_IMAGE019

Figure 992614DEST_PATH_IMAGE040
Figure 992614DEST_PATH_IMAGE040

其中，

Figure 554045DEST_PATH_IMAGE021
为k时刻自动驾驶观光车的后验状态向量，
Figure 259964DEST_PATH_IMAGE022
为卡尔曼增益，
Figure 519038DEST_PATH_IMAGE002
为观测 值，
Figure 977702DEST_PATH_IMAGE023
为k时刻状态变量到观测变量的转换矩阵，
Figure 929608DEST_PATH_IMAGE024
为k时刻后验估计协方差；in,
Figure 554045DEST_PATH_IMAGE021
is the posterior state vector of the self-driving sightseeing car at time k ,
Figure 259964DEST_PATH_IMAGE022
for the Kalman gain,
Figure 519038DEST_PATH_IMAGE002
is the observed value,
Figure 977702DEST_PATH_IMAGE023
is the transformation matrix from the state variable to the observed variable at time k ,
Figure 929608DEST_PATH_IMAGE024
Posteriori estimates covariance for time k ;

其中，

Figure 575353DEST_PATH_IMAGE025
-1； in,
Figure 575353DEST_PATH_IMAGE025
-1 ;

式中，

Figure 247905DEST_PATH_IMAGE026
为服从均值为0正态分布的高斯白噪声； In the formula,
Figure 247905DEST_PATH_IMAGE026
Gaussian white noise that obeys a normal distribution with a mean of 0;

其中，后验状态向量为k时刻自动驾驶观光车状态向量的当前最优估计向量，其中包含自动驾驶观光车当前位置的最优估计值。Among them, the posterior state vector is the current optimal estimation vector of the state vector of the self-driving sightseeing car at time k , which contains the optimal estimated value of the current position of the self-driving sightseeing car.

系统根据上述最优估计值，计算自动驾驶观光车与1号目标点的最近距离，该最近的距离为自动驾驶光观车对第一个目标点的跟踪误差值；The system calculates the shortest distance between the self-driving sightseeing car and the No. 1 target point based on the above optimal estimated value, and the shortest distance is the tracking error value of the first target point by the self-driving sightseeing car;

接着，自动驾驶观光车沿目标路径自动循迹行驶，自适应卡尔曼滤波算法实时对当前行驶位置进行最优估计，并计算最优估计值与下一个目标点，即2号目标点的距离。随着自动驾驶观光车接近并远离2号目标点，自动驾驶观光车与2号目标点的距离先减小后增大，系统自动记录自动驾驶观光车与2号目标点的最小距离，此为自动驾驶观光车对第2个目标点的跟踪误差值。Then, the self-driving sightseeing car automatically tracks along the target path, and the adaptive Kalman filter algorithm makes an optimal estimate of the current driving position in real time, and calculates the distance between the optimal estimate and the next target point, which is the No. 2 target point. As the self-driving sightseeing car approaches and moves away from the No. 2 target point, the distance between the self-driving sightseeing car and the No. 2 target point first decreases and then increases, and the system automatically records the minimum distance between the self-driving sightseeing car and the No. 2 target point, which is The tracking error value of the self-driving sightseeing car to the second target point.

自适应卡尔曼滤波算法对自动驾驶观光车当前位置进行最优估计，计算当前位置与下个目标点，如N号目标点的距离，随着自动驾驶观光车接近并远离N号目标点，自动驾驶观光车与N号目标点的距离先减小后增大，系统记录自动驾驶观光车与N号目标点的最小距离，此为自动驾驶观光车对为第n个坐标点的跟踪误差值。The adaptive Kalman filter algorithm optimally estimates the current position of the self-driving sightseeing car, and calculates the distance between the current position and the next target point, such as the target point N. As the self-driving sightseeing car approaches and moves away from the target point N , it automatically The distance between the driving sightseeing car and the No. N target point first decreases and then increases, and the system records the minimum distance between the self-driving sightseeing car and the No. N target point. This is the tracking error value of the self-driving sightseeing car pair as the nth coordinate point.

系统根据自动驾驶观光车对上述M个坐标点的跟踪误差，即可评估自动驾驶观光车的路径跟踪能力，M个跟踪误差的最大值即为自动驾驶观光车路径跟踪的最大误差值，即其跟踪精度。The system can evaluate the path tracking ability of the self-driving sightseeing car based on the tracking errors of the above-mentioned M coordinate points by the self-driving sightseeing car. The maximum value of the M tracking errors is the maximum error value of the path tracking of the self-driving sightseeing car, that is Tracking accuracy.

上述自适应系数s通过如下公式计算；The above adaptive coefficient s is calculated by the following formula;

Figure 29916DEST_PATH_IMAGE027
Figure 29916DEST_PATH_IMAGE027

其中，N为超参数，表示所取的k时刻之前的时间窗宽度；Among them, N is a hyperparameter, representing the time window width before the selected k moment;

Figure 215041DEST_PATH_IMAGE028
为第k时刻观测向量z k 与预测值
Figure 895552DEST_PATH_IMAGE029
之间的偏差，计算公式如下；
Figure 215041DEST_PATH_IMAGE028
is the observed vector z k and the predicted value at the kth moment
Figure 895552DEST_PATH_IMAGE029
The deviation between is calculated as follows;

Figure 745697DEST_PATH_IMAGE030
Figure 745697DEST_PATH_IMAGE030

Figure 929685DEST_PATH_IMAGE031
为
Figure 31850DEST_PATH_IMAGE028
的转置；
Figure 929685DEST_PATH_IMAGE031
for
Figure 31850DEST_PATH_IMAGE028
the transposition of

Tr()为矩阵迹计算函数，下标j对应由0开始至N-1号自动驾驶观光车驶过的各标记点序号；Tr() is a matrix trace calculation function, and the subscript j corresponds to the serial number of each marker point from 0 to N -1 automatic driving sightseeing car passing by;

当s≥ s 0时，更新步骤S12会增大自动驾驶观光车先验状态向量的偏差，从而增大更新阶段观测信息的权重。When s ≥ s 0 , the update step S12 will increase the deviation of the prior state vector of the self-driving sightseeing car, thereby increasing the weight of the observation information in the update stage.

自动驾驶观光车的状态向量，包括先验状态向量及后验状态向量，均根据自动驾驶观光车的位置、速度和姿态建立状态向量，状态向量包括10个维度，包括3个位置参数，3个速度参数和四元数。The state vector of the self-driving sightseeing car, including the prior state vector and the posterior state vector, is established according to the position, speed and attitude of the self-driving sightseeing car. The state vector includes 10 dimensions, including 3 position parameters, 3 Velocity parameters and quaternions.

以上，RTK高精度定位设备为采用载波相位差分技术的定位设备，GPS/INS组合导航系统为结合了GPS全球定位系统和惯性导航系统的组合系统，其均为现有的市售产品。Above, the RTK high-precision positioning equipment is a positioning equipment using carrier phase difference technology, and the GPS/INS integrated navigation system is a combined system combining GPS global positioning system and inertial navigation system, which are all existing commercially available products.

本实施例的自动驾驶观光车路径跟踪精度检测方法，通过依次获取自动驾驶光观车对各目标点的跟踪误差，并根据所有目标点跟踪误差的最大值，得到自动驾驶光观车的路径跟踪精度。同时，在对各目标点跟踪误差的获取时，通过自适应系数开关函数的设定，实现了卡尔曼滤波算法系数的自适应设定，进而实现了卡尔曼滤波算法状态向量最优估计过程中，预测值和观测值权重的动态自适应调整，克服了递推过程中卡尔曼增益发散的问题。The method for detecting the path tracking accuracy of the self-driving sightseeing car in this embodiment obtains the tracking error of the self-driving sightseeing car on each target point in sequence, and obtains the path tracking of the self-driving sightseeing car according to the maximum value of the tracking errors of all target points precision. At the same time, when obtaining the tracking error of each target point, through the setting of the adaptive coefficient switch function, the adaptive setting of the coefficient of the Kalman filter algorithm is realized, and then the optimal estimation process of the state vector of the Kalman filter algorithm is realized. , the dynamic adaptive adjustment of the predicted value and the weight of the observed value overcomes the problem of Kalman gain divergence in the recursive process.

其中，状态转移矩阵

Figure 668499DEST_PATH_IMAGE034
、控制矩阵
Figure 853493DEST_PATH_IMAGE035
、控制输入向量
Figure 157566DEST_PATH_IMAGE036
由自动驾驶观光车固有 属性确定，
Figure 622177DEST_PATH_IMAGE002
为观测值，转换矩阵
Figure 277280DEST_PATH_IMAGE023
由自动驾驶观光车及观测值的参照系确定，上述各变 量或参数值的具体计算方法为本领域常规技术，此处不再赘述，具体可参考相关文献，如 《卡尔曼滤波及其实时应用》（作者：戴洪德，李娟，戴邵武等，出版社：清华大学出版社，2018 年）。超参数
Figure DEST_PATH_IMAGE041
为预设的自适应系数阈值，其由操作人员根据经验设定。 Among them, the state transition matrix
Figure 668499DEST_PATH_IMAGE034
, control matrix
Figure 853493DEST_PATH_IMAGE035
, control input vector
Figure 157566DEST_PATH_IMAGE036
Determined by the inherent properties of the self-driving sightseeing car,
Figure 622177DEST_PATH_IMAGE002
For observations, the transformation matrix
Figure 277280DEST_PATH_IMAGE023
Determined by the reference system of the self-driving sightseeing car and the observed values, the specific calculation methods of the above-mentioned variables or parameter values are conventional techniques in the field, and will not be repeated here. For details, please refer to relevant literature, such as "Kalman Filter and Its Real-time Application (Authors: Dai Hongde, Li Juan, Dai Shaowu, etc., Publisher: Tsinghua University Press, 2018). hyperparameters
Figure DEST_PATH_IMAGE041
is the preset adaptive coefficient threshold, which is set by the operator based on experience.

以上所述仅是本发明优选的实施方式，用于辅助本领域技术人员实现相应的技术方案，而并不用于限制本发明的保护范围，本发明的保护范围由所附权利要求限定。应当指出，对于本领域的普通技术人员来说，在本发明的技术方案基础上，可做出若干与其等同的改进和变型，这些改进和变型也应视为本发明的保护范围。同时，应当理解，虽然本说明书按照上述实施方式加以描述，但并非每个实施方式仅包含一个独立的技术方案，说明书的这种叙述方式仅仅是为清楚起见，本领域技术人员应当将说明书作为一个整体，各实施例的技术方案也可以经适当组合，形成本领域技术人员可以理解的其他实施方式。The above descriptions are only preferred implementations of the present invention, and are used to assist those skilled in the art to realize corresponding technical solutions, and are not used to limit the protection scope of the present invention, which is defined by the appended claims. It should be pointed out that those skilled in the art can make several equivalent improvements and modifications on the basis of the technical solution of the present invention, and these improvements and modifications should also be regarded as the protection scope of the present invention. At the same time, it should be understood that although this specification is described according to the above-mentioned implementation modes, not each implementation mode only includes an independent technical solution. In general, the technical solutions of the various embodiments can also be properly combined to form other implementations that can be understood by those skilled in the art.

Claims (5)

1. A method for detecting the path tracking precision of an automatic driving sightseeing vehicle comprises the following steps;
collecting coordinate data of a plurality of target points on a planned path;
sequentially acquiring the tracking error of the automatic driving sightseeing vehicle to each target point in the automatic driving process of the automatic driving sightseeing vehicle according to the path;
the path tracking precision of the automatic driving sightseeing vehicle is the maximum error value of the plurality of tracking errors;
the method is characterized in that: the method for acquiring the tracking error of the automatic driving sightseeing vehicle on a certain target point comprises the following steps;
step S1;
acquiring an optimal estimation value of the current position of the automatic driving sightseeing vehicle at preset time intervals in the process that the automatic driving sightseeing vehicle approaches to or leaves away from a target point;
step S2;
calculating the distance between the optimal estimated value of the current position of the automatic driving sightseeing vehicle and a target point according to the optimal estimated value of the current position of the automatic driving sightseeing vehicle obtained in the step S1;
step S3;
the distance between the optimal estimated values obtained in the step S1 and the step S2 and the target point is the smallest distance, namely the tracking error of the automatic driving sightseeing vehicle to the target point;
in step S1, akThe method for acquiring the optimal estimation value of the current position of the automatic driving sightseeing bus at any moment comprises the following steps;
step S11;
establishing a state vector based on state parameters of an autonomous driving sightseeing vehicle
Figure 829029DEST_PATH_IMAGE001
And observation vector
Figure 563767DEST_PATH_IMAGE002
The state vector comprises the position, speed and course angle parameters of the automatic driving sightseeing vehicle, and the state vector
Figure 657756DEST_PATH_IMAGE001
And observation vectors
Figure 395030DEST_PATH_IMAGE002
The expression of (a) is as follows;
Figure 433742DEST_PATH_IMAGE003
+
Figure 683589DEST_PATH_IMAGE004
Figure 671398DEST_PATH_IMAGE005
predictionkMoment autopilot sightseeing vehicle prior state vector
Figure 274680DEST_PATH_IMAGE006
And a priori covariance
Figure 15234DEST_PATH_IMAGE007
The prediction formula is as follows;
Figure 311349DEST_PATH_IMAGE008
Figure 276201DEST_PATH_IMAGE009
wherein,
Figure 729179DEST_PATH_IMAGE010
is from the first tokTime-1 to time-1kA state transition matrix for automatically driving the sightseeing vehicle at all times,
Figure 934026DEST_PATH_IMAGE011
is composed ofk-a posterior state vector of the autonomous driving sightseeing vehicle at time 1,
Figure 384730DEST_PATH_IMAGE012
is composed ofkA control matrix for automatically driving the sightseeing bus at any moment,
Figure 596400DEST_PATH_IMAGE013
is composed ofkControl input vector of time autopilot sightseeing vehicle, T: (s) In order to adapt the coefficient switching function,
Figure 56331DEST_PATH_IMAGE014
is composed ofk-a posteriori estimated covariance of the autonomous driving sightseeing vehicles at time 1,
Figure 709161DEST_PATH_IMAGE015
is composed of
Figure 268449DEST_PATH_IMAGE010
The transpose matrix of (a) is,
Figure 154366DEST_PATH_IMAGE016
is composed ofkTime of day, system noise vector obeying normal distribution with mean 0, where T(s)As an adaptive coefficient switching function, as follows;
T(s)=
Figure 90092DEST_PATH_IMAGE017
in the formulasFor adaptive coefficients, hyper-parameterss 0 An adaptive coefficient threshold greater than 1;
step S12;
updating the prior state vector obtained in step S11 according to the observed value
Figure 394165DEST_PATH_IMAGE006
And a priori skewed estimated variance
Figure 655513DEST_PATH_IMAGE007
The update formula is shown below;
Figure 28726DEST_PATH_IMAGE018
Figure 236984DEST_PATH_IMAGE019
wherein,
Figure 672862DEST_PATH_IMAGE020
is composed ofkThe posterior state vector of the sightseeing vehicle is automatically driven at any moment,
Figure 105112DEST_PATH_IMAGE021
in order to be the basis of the kalman gain,
Figure 231200DEST_PATH_IMAGE002
in order to obtain the observed value, the measured value,
Figure 243149DEST_PATH_IMAGE022
is composed ofkThe transition matrix of the state variable to the observed variable at the moment,
Figure 725077DEST_PATH_IMAGE023
is composed ofkEstimating covariance a posteriori of time;
wherein,
Figure 577496DEST_PATH_IMAGE024
-1 ；
in the formula,
Figure 410454DEST_PATH_IMAGE025
is Gaussian white noise which follows normal distribution with the mean value of 0;
the posterior state vector iskAnd the current optimal estimation vector of the state vector of the automatic driving sightseeing vehicle at the moment comprises the optimal estimation value of the current position of the automatic driving sightseeing vehicle.
2. The automated sightseeing vehicle path tracking accuracy detection method of claim 1, wherein: adaptive coefficients in step S11sCalculated by the following formula;
Figure 209782DEST_PATH_IMAGE026
wherein,Nfor hyper-parameters, indicate what is takenkA time window width before the time of day;
Figure 342955DEST_PATH_IMAGE027
is as followskTime of day observation vector z k And the predicted value
Figure 320269DEST_PATH_IMAGE028
The calculation formula of the deviation is as follows;
Figure 624211DEST_PATH_IMAGE029
Figure 40280DEST_PATH_IMAGE030
is composed of
Figure 762380DEST_PATH_IMAGE027
Transposing;
tr () is a matrix trace calculation function, subscriptjCorrespond to from 0 toN-number 1 of each marker point that the autonomous driving sightseeing vehicle has traveled;
when in uses≥ s 0 In the updating step S12, the deviation of the prior state vector of the autonomous driving sightseeing vehicle is increased, so as to increase the weight of the observation information in the updating stage.
3. The automatic driving sightseeing vehicle path tracking accuracy detection method of claim 1, characterized in that: the target point is collected by RTK high-precision positioning equipment.
4. The automated sightseeing vehicle path tracking accuracy detection method of claim 1, wherein: and a GPS/INS integrated navigation system is installed on the automatic driving sightseeing vehicle, and the observation value is acquired by the GPS/INS integrated navigation system.
5. The automatic driving sightseeing vehicle path tracking accuracy detection method of claim 1, characterized in that: the prior state vector and the posterior state vector include position, speed and attitude parameters of the autonomous driving sightseeing vehicle.
CN202211458202.1A 2022-11-21 2022-11-21 A method for detecting the path tracking accuracy of an autonomous sightseeing vehicle Active CN115790629B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202211458202.1A CN115790629B (en) 2022-11-21 2022-11-21 A method for detecting the path tracking accuracy of an autonomous sightseeing vehicle

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202211458202.1A CN115790629B (en) 2022-11-21 2022-11-21 A method for detecting the path tracking accuracy of an autonomous sightseeing vehicle

Publications (2)

Publication Number Publication Date
CN115790629A true CN115790629A (en) 2023-03-14
CN115790629B CN115790629B (en) 2025-06-27

Family

ID=85439522

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202211458202.1A Active CN115790629B (en) 2022-11-21 2022-11-21 A method for detecting the path tracking accuracy of an autonomous sightseeing vehicle

Country Status (1)

Country Link
CN (1) CN115790629B (en)

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN116047567A (en) * 2023-04-03 2023-05-02 长沙金维信息技术有限公司 Deep learning assistance-based guard and inertial navigation combined positioning method and navigation method

Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP2010203799A (en) * 2009-02-27 2010-09-16 Toshiba Corp Target tracking system
CN108646277A (en) * 2018-05-03 2018-10-12 山东省计算中心（国家超级计算济南中心） The Beidou navigation method adaptively merged with Extended Kalman filter based on robust
CN110455287A (en) * 2019-07-24 2019-11-15 南京理工大学 Adaptive Unscented Kalman Particle Filter Method
CN110954132A (en) * 2019-10-31 2020-04-03 太原理工大学 A Navigation Fault Identification Method Using GRNN Assisted Adaptive Kalman Filtering
CN114488224A (en) * 2021-12-24 2022-05-13 西南交通大学 An Adaptive Filtering Method for Satellite Centralized Autonomous Navigation

Patent Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP2010203799A (en) * 2009-02-27 2010-09-16 Toshiba Corp Target tracking system
CN108646277A (en) * 2018-05-03 2018-10-12 山东省计算中心（国家超级计算济南中心） The Beidou navigation method adaptively merged with Extended Kalman filter based on robust
CN110455287A (en) * 2019-07-24 2019-11-15 南京理工大学 Adaptive Unscented Kalman Particle Filter Method
CN110954132A (en) * 2019-10-31 2020-04-03 太原理工大学 A Navigation Fault Identification Method Using GRNN Assisted Adaptive Kalman Filtering
CN114488224A (en) * 2021-12-24 2022-05-13 西南交通大学 An Adaptive Filtering Method for Satellite Centralized Autonomous Navigation

Non-Patent Citations (1)

* Cited by examiner, † Cited by third party
Title
苏天祥;文援兰;朱俊;: "双自适应因子滤波算法", 测绘学报, no. 01, pages 26 - 31 *

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN116047567A (en) * 2023-04-03 2023-05-02 长沙金维信息技术有限公司 Deep learning assistance-based guard and inertial navigation combined positioning method and navigation method

Also Published As

Publication number Publication date
CN115790629B (en) 2025-06-27

Similar Documents

Publication Publication Date Title
CN108614258B (en) An underwater positioning method based on single underwater acoustic beacon distance measurement
CN101464152B (en) Adaptive filtering method for SINS/GPS combined navigation system
CN102508278B (en) Adaptive filtering method based on observation noise covariance matrix estimation
CN105741546A (en) Intelligent vehicle target tracking system through integration of road side equipment and vehicle sensor and method thereof
WO2018072350A1 (en) Vehicle trajectory prediction method and device
CN109343095B (en) Vehicle-mounted navigation vehicle combined positioning device and combined positioning method thereof
JP2024103654A (en) Measurement accuracy calculation device, self-position estimation device, control method, program, and storage medium
CN105809126A (en) Intelligent vehicle target tracking system and method employing DSRC and vehicle sensor in combination
CN110031876A (en) A kind of vehicle mounted guidance tracing point offset antidote based on Kalman filtering
CN109115225A (en) A kind of unmanned operation grain combine air navigation aid and navigation device
CN103395435B (en) A kind of high-precision high-speed train real-time positioning system method
CN114440892B (en) Self-positioning method based on topological map and odometer
JP2024105508A (en) Output device, control method, program, and storage medium
CN106643724A (en) Method for particle filter indoor positioning based on map information and position self-adaption correction
JP2024161130A (en) Self-location estimation device, control method, program, and storage medium
JP2008077349A (en) Vehicle state quantity estimation device and vehicle steering control device using the device
CN113280821B (en) Underwater multi-target tracking method based on slope constraints and backtracking search
CN114777769A (en) Outdoor substation inspection robot positioning system and method thereof
CN115790629A (en) A detection method for the path tracking accuracy of an autonomous sightseeing vehicle
CN112558072A (en) Vehicle positioning method, device, system, electronic equipment and storage medium
CN110243363B (en) A real-time positioning method of AGV based on the combination of low-cost IMU and RFID technology
CN114035154B (en) Single-station radio frequency signal positioning method assisted by motion parameters
Holzapfel et al. Road profile recognition for autonomous car navigation and Navstar GPS support
CN112666519B (en) A high-precision positioning method for underwater targets based on generalized second-order delay difference
CN111307136B (en) Underwater navigation terrain matching navigation method for double intelligent underwater robots

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
SE01 Entry into force of request for substantive examination
SE01 Entry into force of request for substantive examination
GR01 Patent grant
GR01 Patent grant