GNSS data fault simulation method and device, electronic equipment and storage medium

Abstract

Embodiments of the present application provide a GNSS data fault simulation method and device, electronic equipment and storage medium, relating to the technical field of satellite navigation and positioning. The method comprises: obtaining original GNSS road test data under a target scene and performing RTK positioning solution to obtain a plurality of feature data, constructing a decision tree classification model based on each feature data, identifying and extracting a target scene epoch interval in the original GNSS road test data through the decision tree classification model, generating a plurality of fault models based on the target scene epoch interval and a plurality of fault data in the interval, determining a plurality of fault inclusion criteria under the target scene to obtain fault data through the corresponding fault model according to each fault inclusion criterion, and injecting the fault data into the corresponding original GNSS road test data to obtain GNSS fault simulation data. Embodiments of the present application can quickly and accurately obtain GNSS fault simulation data, meeting the data quantity requirement of autonomous driving integrity under the target scene.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of satellite navigation and positioning, in particular to a GNSS data fault simulation method and device, an electronic device and a storage medium. BACKGROUND

[0002] In addition to providing navigation functions such as positioning, speed measurement and time service, the global navigation satellite system also needs to take certain mechanisms to ensure the reliability of the above functions considering the possibility of abnormal conditions in each part of the system. Especially when GNSS positioning is applied to car navigation and positioning, the safety problem needs to be considered.

[0003] Based on this, the concept of global navigation satellite system integrity is introduced, which is defined as the ability to warn users within a given time when the system cannot provide navigation information due to a fault. Therefore, the probability of system issuing dangerous misleading information is an important indicator of the integrity algorithm. In the field of autonomous driving, this probability is HMI 1e-6 / h in the high-speed scenario, i.e. one error every 1000000 hours (i.e. 114 years). Since it is difficult to collect observation data of the above order of magnitude in actual road testing, it is necessary to simulate GNSS observation data collected by the receiver in the high-speed scenario to simulate road test data exceeding 1000000 hours to serve the subsequent GNSS positioning integrity algorithm. SUMMARY

[0004] Therefore, the purpose of the present application is to provide a GNSS data fault simulation method, device, electronic device and storage medium to quickly and accurately obtain GNSS fault simulation data, meeting the data quantity requirement of autonomous driving integrity in the target scenario.

[0005] To achieve the above purpose, the technical solutions adopted by the embodiments of the present application are as follows:

[0006] In a first aspect, the embodiments of the present application provide a GNSS data fault simulation method applied to an electronic device.

[0007] The method comprises:

[0008] Obtaining original GNSS road test data in a target scenario and performing RTK positioning solution to obtain a plurality of feature data;

[0009] Constructing a decision tree classification model based on each feature data;

[0010] Identifying and extracting a target scenario epoch interval in the original GNSS road test data through the decision tree classification model;

[0011] generate a plurality of fault models under the target scene based on the target scene epoch interval and fault data in the target scene epoch interval;

[0012] determine a plurality of fault inclusion criteria under the target scene, so as to obtain fault data under the target scene according to the corresponding fault model through each fault inclusion criterion;

[0013] inject the fault data under the target scene into the corresponding original GNSS road test data to obtain GNSS simulation data.

[0014] Optionally, the step of obtaining original GNSS road test data under the target scene and performing RTK positioning calculation to obtain a plurality of feature data comprises:

[0015] obtaining original GNSS road test data under the target scene and performing RTK positioning calculation to obtain PDOP data, number of pseudorange observation values, number of phase observation values, number of satellites, number of cycle slip satellites and average multipath data;

[0016] the PDOP data, the number of pseudorange observation values, the number of phase observation values, the number of satellites, the number of cycle slip satellites and the average multipath data are taken as feature data.

[0017] Optionally, the step of constructing a decision tree classification model based on each feature data comprises:

[0018] calculating the information gain corresponding to each feature data from the root node;

[0019] taking the feature data corresponding to the maximum information gain as node division feature data;

[0020] establishing a sub-node based on the node division feature data;

[0021] recursively operating on the sub-node to construct a decision tree classification model.

[0022] Optionally, the step of identifying and extracting the target scene epoch interval in the original GNSS road test data through the decision tree classification model comprises:

[0023] obtaining the identification interval of each feature data under the target scene through the decision tree classification model, so as to identify and extract the target scene epoch interval in the original GNSS road test data according to the identification interval of each feature data.

[0024] Optionally, the step of generating a plurality of fault models under the target scene based on the target scene epoch interval and fault data in the target scene epoch interval comprises:

[0025] determine, based on the target scenario epoch interval, cycle slip data and multipath data under the target scenario from the original GNSS road test data;

[0026] perform multiple distribution fittings on the cycle slip data and the multipath data respectively, wherein each of the distribution fittings corresponds to a distribution function;

[0027] calculate a chi-square test quantity of each of the distribution fittings based on the cycle slip data and the distribution function corresponding to each of the distribution fittings, and take the distribution function corresponding to the smallest chi-square test quantity as a cycle slip data failure model function under the target scenario to generate a cycle slip data failure model under the target scenario;

[0028] calculate a chi-square test quantity of each of the distribution fittings based on the multipath data and the distribution function corresponding to each of the distribution fittings, and take the distribution function corresponding to the smallest chi-square test quantity as a multipath data failure model function under the target scenario to generate a multipath data failure model under the target scenario;

[0029] take the cycle slip data failure model and the multipath data failure model under the target scenario as the failure model under the target scenario.

[0030] Optionally, the step of determining a plurality of failure inclusion criteria under the target scenario to obtain failure data under the target scenario according to each of the failure inclusion criteria through the corresponding failure model comprises:

[0031] determine a cycle slip occurrence probability, a multipath occurrence failure judgment threshold and an ephemeris failure occurrence probability under the target scenario, and establish the failure inclusion criteria under the target scenario according to the cycle slip occurrence probability, the multipath occurrence failure judgment threshold and the ephemeris failure occurrence probability under the target scenario;

[0032] obtain cycle slip failure data under the target scenario according to the cycle slip occurrence probability and the cycle slip failure model function;

[0033] obtain multipath failure data under the target scenario through the multipath data failure model function under the target scenario based on the multipath occurrence failure judgment threshold under the target scenario;

[0034] obtain ephemeris failure data under the target scenario through the ephemeris failure model function under the target scenario based on the ephemeris failure occurrence probability under the target scenario;

[0035] take the cycle slip failure data, the multipath failure data and the ephemeris failure data under the target scenario as the failure data under the target scenario.

[0036] Optionally, the original GNSS road test data comprises pseudo-range observation values, phase observation values and ephemeris data.

[0037] The step of injecting the fault data under the target scenario into the corresponding original GNSS road test data to obtain GNSS fault simulation data comprises:

[0038] The fault data under the target scenario is injected into the pseudo-range observation values, phase observation values and ephemeris data to obtain GNSS fault simulation data.

[0039] In a second aspect, an embodiment of the present application provides a GNSS data fault simulation device applied to an electronic device.

[0040] The device comprises:

[0041] A solving module is configured to obtain original GNSS road test data under a target scenario and perform RTK positioning solving to obtain a plurality of feature data.

[0042] A model generating module is configured to construct a decision tree classification model based on the feature data, identify and extract a target scenario epoch interval in the original GNSS road test data through the decision tree classification model, and generate a plurality of fault models under the target scenario based on the target scenario epoch interval and fault data in the target scenario epoch interval.

[0043] A simulation module is configured to determine a plurality of fault inclusion criteria under the target scenario, obtain fault data under the target scenario through the corresponding fault models according to the fault inclusion criteria, and inject the fault data under the target scenario into the corresponding original GNSS road test data to obtain GNSS simulation data.

[0044] In a third aspect, an embodiment of the present application provides an electronic device comprising a memory and a processor.

[0045] The memory is configured to store a computer program.

[0046] The processor is configured to execute the computer program to implement the GNSS data fault simulation method provided in the above-mentioned first aspect embodiment and / or possible implementation manners in combination with the above-mentioned first aspect embodiment.

[0047] In a fourth aspect, an embodiment of the present application provides a computer readable storage medium having a computer program stored thereon, and the computer program is executed by a processor to implement the GNSS data fault simulation method provided in the above-mentioned first aspect embodiment and / or possible implementation manners in combination with the above-mentioned first aspect embodiment.

[0048] The beneficial effects of the embodiments of the present application include, for example:

[0049] The GNSS data fault simulation method, device, electronic equipment and storage medium provided by the embodiment of the present application obtain a plurality of feature data by obtaining original GNSS road test data under a target scene (for example, a high-speed scene) and performing RTK positioning calculation, construct a decision tree classification model based on each feature data, identify and extract a target scene epoch interval in the original GNSS road test data through the decision tree classification model, then generate a plurality of fault models under the target scene based on the target scene epoch interval and fault data in the target scene epoch interval, then perform fault injection work according to the fault model and a plurality of fault inclusion standards determined, and finally obtain accurate GNSS fault simulation data, which meets the data quantity requirement of the automatic driving integrity under the target scene, thereby serving the GNSS positioning integrity algorithm.

[0050] In order to make the above objectives, characteristics and advantages of the present application more apparent, the following preferred embodiments are specifically described below, and the accompanying drawings are referred to for detailed description. BRIEF DESCRIPTION OF DRAWINGS

[0051] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the following will briefly introduce the drawings needed to be used in the embodiments. It should be understood that the following drawings only show some embodiments of the present application, and therefore should not be regarded as a limitation on the scope, and for those skilled in the art, other related drawings can also be obtained without creative labor on the basis of these drawings.

[0052] Figure 1 An exemplary structural block diagram of an electronic equipment provided by the embodiment of the present application is shown;

[0053] Figure 2 A flowchart of a GNSS data fault simulation method provided by the embodiment of the present application is shown;

[0054] Figure 3 A flowchart of a GNSS data fault simulation method provided by the embodiment of the present application is shown;

[0055] Figure 4 A flowchart of a GNSS data fault simulation method provided by the embodiment of the present application is shown;

[0056] Figure 5 A flowchart of a GNSS data fault simulation method provided by the embodiment of the present application is shown;

[0057] Figure 6 A flowchart of a GNSS data fault simulation method provided by the embodiment of the present application is shown;

[0058] Figure 7A flowchart of a GNSS data fault simulation method provided by an embodiment of the present application is shown Figure Six ;

[0059] Figure 8 A flowchart of a GNSS data fault simulation method provided by an embodiment of the present application is shown Figure Seven ;

[0060] Figure 9 An exemplary structural block diagram of a GNSS data fault simulation device provided by an embodiment of the present application is shown.

[0061] Icon: 110 - electronic device; 1101 - memory; 1102 - processor; 1103 - communication interface; 300 - GNSS data fault simulation device; 301 - solving module; 302 - model generation module; 303 - simulation module. DETAILED DESCRIPTION

[0062] The technical solutions in the embodiments of the present application will be described clearly and completely below with reference to the accompanying drawings in the embodiments of the present application. Obviously, the described embodiments are only some of the embodiments of the present application, but not all the embodiments. The components of the embodiments of the present application described and shown in the accompanying drawings can be arranged and designed in various different configurations.

[0063] Therefore, the detailed description of the embodiments of the present application provided in the accompanying drawings below is not intended to limit the scope of the claimed present application, but only represents selected embodiments of the present application. All other embodiments obtained by those skilled in the art based on the embodiments of the present application without creative labor are within the scope of protection of the present application.

[0064] It should be noted that the relational terms such as first and second and the like are used only to differentiate one entity or action from another, and do not necessarily require or imply that these entities or actions occur in any such actual relationship or order. Moreover, the terms "comprising", "including" or any other variant thereof are intended to cover non-exclusive inclusion, so that a process, method, article or apparatus that includes a list of elements not only includes those elements, but also includes other elements not expressly listed or inherent to such process, method, article or apparatus. Without more limitations, the element defined by the statement "comprising a" does not exclude the presence of additional identical elements in the process, method, article or apparatus that includes the element.

[0065] It should be noted that the features in the embodiments of the present application can be combined with each other without conflict.

[0066] There are mainly two kinds of technologies that can provide high-precision position by GNSS (Global Navigation Satellite System): differential positioning (RTK) and precise point positioning (PPP). Among them, the RTK positioning technology requires the receiver to receive and process satellite signals, as well as differential information, to eliminate atmospheric errors such as ionosphere / troposphere and satellite orbit / satellite clock errors through difference, realize ambiguity fixing of double difference, and then obtain high-precision position; the PPP positioning technology requires processing satellite signals, as well as precise satellite orbit clock error / pseudorange carrier bias / atmosphere correction information provided by service providers, to correct and fix ambiguity, and then obtain high-precision position.

[0067] In addition to providing navigation functions such as positioning, speed measurement, and time service, the global navigation satellite system also needs to take certain mechanisms to ensure the reliability of the above functions considering the possibility of abnormal conditions in each part of the system. Especially when GNSS positioning is applied to car navigation positioning, the safety problem needs to be considered.

[0068] Based on this, the concept of global navigation satellite system integrity is introduced, which is defined as the ability to warn users within a given time when the system cannot provide navigation information due to failure. Therefore, the probability of system issuing dangerous misleading information is an important indicator of integrity algorithm. In the field of autonomous driving, this probability is 1e-6 / h in high-speed scenarios, i.e., one error every 1000000 hours (i.e., 114 years). Since it is difficult to collect a large number of observation data in actual road testing, GNSS observation data collected by the receiver in high-speed scenarios need to be simulated for failure to simulate more than 1000000 hours of road testing data to serve subsequent GNSS positioning integrity algorithms.

[0069] The traditional GNSS observation data failure simulation method usually includes the following steps:

[0070] Step one, extracting errors related to GNSS receivers from GNSS receiver observation data, wherein the errors related to GNSS receivers include at least one of pseudorange multipath error, carrier multipath error, carrier cycle slip error, and Doppler frequency shift error;

[0071] Step two, analyzing and modeling the failure characteristics of the errors related to the GNSS receiver, and generating a failure model;

[0072] Step three, injecting simulated failures related to GNSS receivers into the GNSS receiver observation data to be injected according to the failure model, to obtain GNSS receiver failure simulation data.

[0073] The GNSS observation data fault simulation method can model and simulate errors related to the GNSS receiver, and can generate any number of observation data for GNSS positioning algorithm upgrade iteration and integrity verification. However, there are still defects, for example, the GNSS observation data fault simulation method is only suitable for GNSS receiver fault simulation in general scenarios, cannot effectively analyze and extract GNSS data fault models in high-speed scenarios, and only extracts and models errors of observation data of the GNSS receiver, without fully considering satellite ephemeris errors, and without describing mathematical models of multipath and cycle slip faults in high-speed scenarios.

[0074] Based on this, the embodiment of the present application provides a GNSS data fault simulation method to solve the above problems.

[0075] Please refer to Figure 1 , Figure 1 An exemplary structural block diagram of an electronic device 110 provided by the embodiment of the present application is shown, which can be a PC or other intelligent terminal for GNSS data fault simulation, as shown in Figure 1 The electronic device 110 includes a memory 1101, a processor 1102 and a communication interface 1103, which are directly or indirectly electrically connected to each other to realize data transmission or interaction. For example, these elements can be electrically connected to each other through one or more communication buses or signal lines.

[0076] The memory 1101 can be used to store software programs and modules, and the processor 1102 can execute various functions and data processing by executing the software programs and modules stored in the memory 1101. The communication interface 1103 can be used for signaling or data communication with other node devices.

[0077] The memory 1101 can be, but is not limited to, a random access memory (RAM), a read-only memory (ROM), a programmable read-only memory (PROM), an erasable programmable read-only memory (EPROM), an electrically erasable programmable read-only memory (EEPROM) and the like.

[0078] The processor 1102 can be an integrated circuit chip having a signal processing capability. The processor 1102 can be a general-purpose processor, including a central processing unit (CPU), a network processor (NP), etc.; or can be a digital signal processor (DSP), an application specific integrated circuit (ASIC), a field-programmable gate array (FPGA) or other programmable logic device, a discrete gate or transistor logic device, a discrete hardware component.

[0079] Based on the above-mentioned electronic device 110, the GNSS data fault simulation method provided by the embodiment of the present application will be exemplarily described below with the electronic device 110 as the execution subject. Please refer to Figure 2 , Figure 2 The flowchart of the GNSS data fault simulation method provided by the embodiment of the present application is shown.

[0080] As shown in Figure 2 , the GNSS data fault simulation method is applied to the electronic device 110, and the GNSS data fault simulation method can include the following steps:

[0081] S210, obtaining original GNSS road test data under a target scene and performing RTK positioning solution to obtain a plurality of feature data.

[0082] S220, constructing a decision tree classification model based on each feature data.

[0083] S230, identifying and extracting a target scene epoch interval in the original GNSS road test data through the decision tree classification model.

[0084] S240, generating a plurality of fault models under the target scene based on the target scene epoch interval and the fault data in the target scene epoch interval.

[0085] S250, determining a plurality of fault inclusion criteria under the target scene to obtain fault data under the target scene through the corresponding fault model according to each fault inclusion criterion.

[0086] S260, injecting the fault data under the target scene into the corresponding original GNSS road test data to obtain GNSS simulation data.

[0087] The above steps achieve the process of obtaining original GNSS road test data under a target scene and performing RTK positioning calculation to obtain a plurality of feature data, constructing a decision tree classification model based on each feature data, identifying and extracting a target scene epoch interval in the original GNSS road test data through the decision tree classification model, then generating a plurality of fault models under the target scene based on the target scene epoch interval and the original GNSS road test data, and then performing fault injection work according to the fault model and a plurality of fault inclusion standards determined, and finally obtaining GNSS fault simulation data.

[0088] In step S210, original GNSS road test data under a target scene is obtained and RTK positioning calculation is performed to obtain a plurality of feature data.

[0089] It should be noted that the entire process of RTK positioning calculation includes data preprocessing, Kalman floating point filter updating, wide item ambiguity fixing, ambiguity fixing and the like, and a plurality of feature data can be obtained in the above process of RTK positioning calculation, for example, a plurality of feature data shown in Table 1 can be obtained:

[0090] Table 1

[0091]

[0092] Based on the above settings, the plurality of feature data obtained in step S210 can be PDOP data, number of pseudorange observations, number of phase observations, number of satellites, number of cycle slip satellites and average multipath data.

[0093] In the embodiment of the application, after obtaining the plurality of feature data, steps S220 and S230 are continued, a decision tree classification model is constructed based on each feature data, and then the target scene epoch interval in the original GNSS road test data is identified and extracted through the decision tree classification model.

[0094] Specifically, the purpose of constructing a decision tree classification model based on each feature data is to use the decision tree classification model to identify and extract the target scene epoch interval in any number of actually obtained original GNSS road test data, that is, to obtain the identification interval of each feature data under the target scene through the constructed decision tree classification model, so as to identify and extract the target scene epoch interval in the original GNSS road test data according to the identification interval of each feature data for subsequent determination of faults under the target scene (i.e., determination of cycle slip data and multipath data under the target scene) from the above original GNSS road test data.

[0095] It should be noted that the decision tree is a non-parametric supervised learning method that can summarize decision rules from a series of data with features and labels, and present these rules in a tree structure to solve classification and regression problems. Decision tree algorithm is easy to understand, suitable for various data, and has good performance in solving various problems. Various ensemble algorithms centered on tree models have been widely used in various industries and fields.

[0096] The decision tree classification model is a machine learning method for solving classification problems. For a classification decision tree, the most important step is to select a few important attributes from a large number of feature value attributes to establish a tree. Each decision tree has many nodes, including root nodes (no incoming edge, outgoing edge, containing initial judgment on features), intermediate nodes (having incoming and outgoing edges, the incoming edge has only one, the outgoing edge can have many, containing judgment on feature values), and leaf nodes (having incoming edge and no outgoing edge, each leaf node is a class label).

[0097] Further, in a classification problem, the process of classifying instances based on features can be considered as a collection of if-else statements, or as a conditional probability distribution defined on the feature space and the class space. A decision tree usually has three steps: feature selection, decision tree generation, and decision tree pruning. The construction of a decision tree generally starts from the root node, tests a feature of an instance, and assigns the instance to its child node according to the test result. At this time, each child node corresponds to a value of the feature. Such recursive testing and assignment of instances are performed until the leaf node is reached, and finally the instance is classified into the class of the leaf node. The goal of decision tree learning is to construct a decision tree model from a given training data set, so that it can correctly classify instances. The essence of decision tree learning is to induce a set of classification rules from the training set, or to estimate a conditional probability model from the training data set. In order to determine the decisive feature of the current feature value data set, in order to obtain the decisive feature, each feature must be evaluated. After the test is completed, the original data set is divided into several data subsets, which are distributed on all branches of the first decision point. If the data under a branch belong to the same type, the current unordered feature value has correctly classified the data, and there is no need to further divide the data set. If they do not belong to the same type, the data subsets need to be repeatedly divided until all data of the same type are in a data subset.

[0098] Therefore, using the constructed decision tree classification model to identify and extract target scenario epoch intervals from massive original GNSS road test data has high efficiency and is simple and convenient overall compared with other classification models.

[0099] Further, the specific steps of constructing the decision tree classification model based on each feature data in step S220 can first calculate the information gain corresponding to each feature data from the root node, take the feature data corresponding to the maximum information gain as the node division feature data, then establish the sub-node based on the node division feature data, and finally perform recursive operation on the sub-node to construct the decision tree classification model. In the process of construction, if there is no feature data to be selected or the categories of the feature data are completely the same, the final decision tree classification model can be output.

[0100] Further, the original GNSS road test data also needs to be input into the finally obtained decision tree classification model, so that the decision tree classification model outputs the classification result, and the classification result is compared with the actual classification result. If the accuracy meets the preset requirement, the identification interval of each feature data in the target scene can be output.

[0101] For example, based on the plurality of feature data in Table 1 in the foregoing, if the target scene is set as a high-speed scene, the identification interval of each feature data in the high-speed scene obtained through the decision tree classification model can be respectively: PDOP<=1.3, number of pseudorange observation values>=35, number of phase observation values>=35, number of satellites>=28, number of cycle slip satellites<=5, average multipath MP1<=0.66m, MP2<=0.68m. Based on the identification interval of each feature data in the above target scene, the target scene epoch interval in the original GNSS road test data can be identified and extracted according to the identification interval of each feature data.

[0102] In the embodiment of the application, after the target scene epoch interval in the original GNSS road test data is identified and extracted through the decision tree classification model, step S240 is continued, and the fault model in a plurality of target scenes is generated based on the target scene epoch interval and the original GNSS road test data.

[0103] Specifically, step S240 can be implemented by determining the cycle slip data and the multipath data in the target scene based on the target scene epoch interval and the fault data in the target scene epoch interval, then performing a plurality of distribution fittings on the cycle slip data and the multipath data respectively to finally obtain the cycle slip data fault model function in the target scene and the multipath data fault model function in the target scene respectively.

[0104] Exemplarily, the process of performing multiple distribution fitting on the cycle slip data and multipath data in the above steps can respectively perform ten scenario distribution fitting on the cycle slip data and multipath data, and the ten distribution scenarios can be: beta distribution, expon distribution, gamma distribution, lognorm distribution, norm distribution, pearson3 distribution, triang distribution, uniform distribution, weibull_min distribution, and weibull_max distribution.

[0105] Furthermore, while performing multiple distribution fittings on the cycle slip data and multipath data respectively, a chi-square test is also required, that is, the chi-square test amounts of various distribution fittings are calculated based on the cycle slip data and the distribution functions corresponding to the various distribution fittings, and the distribution function corresponding to the minimum chi-square test amount is used as the cycle slip data fault model function under the target scenario to generate the cycle slip data fault model under the target scenario.

[0106] Similarly, it is also necessary to calculate the chi-square test values ​​of various distribution fits based on the multi-path data and the distribution functions corresponding to various distribution fits, and use the distribution function corresponding to the minimum chi-square test value as the multi-path data fault model function under the target scenario to generate the multi-path data fault model under the target scenario.

[0107] It should be noted that the chi-square test can be expressed by the following formula:

[0108]

[0109] Among them, n is the total number of samples, k is the number of equal parts of the sample data in ascending order, and f i is the actual frequency of sample data in each copy, p i The theoretical probability is determined based on the distribution function. For each distribution function, we can first assume that the distribution function of the overall data is the distribution function corresponding to the distribution fit, and then calculate the chi-square test value using the above formula 1, and then use χ 2 The chi-square test method is used to test whether the sample data obeys the hypothesized distribution, that is, the chi-square test value χ 2 The distribution function corresponding to the smallest chi-square test value is used as the final distribution function.

[0110] Therefore, in an embodiment of the present invention, the above-mentioned sample data is cycle slip data and multipath data. If the cycle slip data is divided into 50 parts in ascending order, k is 50. Based on the cycle slip data and the distribution functions corresponding to various distribution fits, the chi-square test values ​​of various distribution fits are calculated by formula 1, and then all the obtained chi-square test values ​​are compared. The distribution function corresponding to the smallest chi-square test value is finally used as the cycle slip data fault model function in the target scenario.

[0111] Similarly, the chi-square test quantity of each distribution fitting is calculated by formula one based on the multi-path data and the corresponding distribution function of each distribution fitting, and then all obtained chi-square test quantities are compared, and the distribution function corresponding to the smallest chi-square test quantity is finally taken as the multi-path data fault model function under the target scene. Finally, the corresponding cycle slip data fault model and multi-path data fault model under the target scene (i.e. the fault models under multiple target scenes) can be obtained.

[0112] In the embodiments of the present application, after obtaining the fault models under multiple target scenes, steps S250 and S260 are continued to determine the multiple fault inclusion criteria under the target scene, so as to obtain the fault data under the target scene according to the corresponding fault model according to each fault inclusion criterion, and then the fault data under the target scene is injected into the corresponding original GNSS road test data to obtain GNSS fault simulation data.

[0113] Specifically, the multiple fault inclusion criteria in step S250 can be specifically the cycle slip occurrence probability under the target scene, the multi-path occurrence fault judgment threshold and the ephemeris fault occurrence probability.

[0114] Further, the cycle slip occurrence probability under the target scene can be obtained by the number of epochs with cycle slip / total number of epochs; the multi-path occurrence fault judgment threshold can be obtained by the lowest multi-path data of the non-fixed solution under the target scene; and the ephemeris fault occurrence probability is generally 1e-5 (1 satellite) and 1e-8 (2 satellites and above).

[0115] Therefore, based on the multiple fault inclusion criteria obtained above, the fault data under the target scene can be obtained according to the corresponding fault model according to each fault inclusion criterion.

[0116] For example, if the fault models under multiple target scenes generated in step S240 are cycle slip data fault model function and multi-path data fault model function, and the cycle slip data fault model function is an expon distribution function, the multi-path data fault model function is a beta distribution function, and the fault model generally selected for ephemeris fault is a ramp fault model, then the model functions corresponding to each fault model finally obtained can be expressed as follows:

[0117] f(x)=exp(-x)

[0118] (Formula two)

[0119]

[0120] f(t)=Au(t-t0)

[0121] (Formula four)

[0122] wherein, formula two is an expon distribution function (i.e., a cycle slip data failure model function under a target scene), formula three is a beta distribution function (i.e., a multipath data failure model function under a target scene), formula four is a model function corresponding to a ramp failure model (i.e., an ephemeris failure model function under a target scene), and u(t) in formula four is a unit step function, t0 is a failure start time, and A is a failure amplitude.

[0123] Therefore, the failure data under the target scene can be obtained according to the corresponding failure model function according to the failure inclusion criteria.

[0124] For example, based on the failure inclusion criteria set in the foregoing and the cycle slip data failure model function, the multipath data failure model function and the ephemeris failure model function obtained, the cycle slip failure data under the target scene can be obtained according to the cycle slip occurrence probability under the target scene through the cycle slip data failure model function under the target scene; the multipath failure data under the target scene can be obtained according to the multipath occurrence failure judgment threshold under the target scene through the multipath data failure model function under the target scene; and the ephemeris failure data under the target scene can be obtained according to the ephemeris failure occurrence probability under the target scene through the ephemeris failure model function under the target scene.

[0125] In the embodiment of the application, based on the steps in the foregoing, the failure data under the target scene can be injected into the corresponding original GNSS road test data in step S260 to obtain GNSS failure simulation data. For example, based on the plurality of failure data obtained in the foregoing, which are cycle slip failure data, multipath failure data and ephemeris failure data, the above failure data under the target scene can be injected into pseudorange observation values, phase observation values and ephemeris data, and then converted into an rtcm file to finally obtain GNSS failure simulation data.

[0126] The GNSS data failure simulation method and device, the electronic equipment and the storage medium provided in the embodiment of the application obtain a plurality of feature data by acquiring original GNSS road test data under a target scene (for example, a high-speed scene) and performing RTK positioning calculation, construct a decision tree classification model based on the feature data, identify and extract a target scene epoch interval in the original GNSS road test data through the decision tree classification model, generate a plurality of failure models under the target scene based on the target scene epoch interval and failure data in the target scene epoch interval, and then perform failure injection work according to the failure models and the plurality of failure inclusion criteria determined, to finally obtain accurate GNSS failure simulation data, meet the data quantity requirement of the autonomous driving integrity under the target scene, and serve the GNSS positioning integrity algorithm.

[0127] Optionally, the specific process of acquiring original GNSS road test data under a target scene and performing RTK positioning calculation to obtain a plurality of feature data can be implemented through the following steps:

[0128] On the basis of Figure 2 , refer to Figure 3 , Figure 3 Fig. 2 shows a flowchart of a GNSS data fault simulation method provided by an embodiment of the present application, and the step S210 of acquiring original GNSS road test data under a target scene and performing RTK positioning calculation to obtain a plurality of feature data, including:

[0129] S211, acquiring original GNSS road test data under a target scene and performing RTK positioning calculation to obtain PDOP data, number of pseudorange observation values, number of phase observation values, number of satellites, number of cycle slip satellites and average multipath data.

[0130] S212, taking the PDOP data, the number of pseudorange observation values, the number of phase observation values, the number of satellites, the number of cycle slip satellites and the average multipath data as feature data.

[0131] The above steps implement the process of acquiring original GNSS road test data under a target scene and performing RTK positioning calculation to obtain a plurality of feature data.

[0132] Optionally, the specific process of constructing a decision tree classification model based on each feature data can be implemented through the following steps:

[0133] On the basis of Figure 2 , refer to Figure 4 , Figure 4 Fig. 3 shows a flowchart of a GNSS data fault simulation method provided by an embodiment of the present application, and the step S220 of constructing a decision tree classification model based on each feature data, including:

[0134] S221, calculating information gain corresponding to each feature data from a root node.

[0135] S222, taking the feature data corresponding to the maximum information gain as node division feature data.

[0136] S223, establishing a subnode based on the node division feature data.

[0137] S224, performing a recursive operation on the subnode to construct a decision tree classification model.

[0138] The above steps implement the process of constructing a decision tree classification model based on each feature data.

[0139] Optionally, the specific process of identifying and extracting the target scenario epoch interval in the original GNSS road test data through the decision tree classification model can be implemented through the following steps:

[0140] On the basis of Figure 2 , refer to Figure 5 , Figure 5 Fig. 4 shows a flowchart of the GNSS data fault simulation method provided by the embodiment of the present application, and the step of identifying and extracting the target scenario epoch interval in the original GNSS road test data through the decision tree classification model in the S230 step, which includes:

[0141] S231, obtaining the identification interval of each feature data under the target scenario through the decision tree classification model, so as to identify and extract the target scenario epoch interval in the original GNSS road test data according to the identification interval of each feature data.

[0142] The above steps realize the process of identifying and extracting the target scenario epoch interval in the original GNSS road test data through the decision tree classification model.

[0143] Optionally, the specific process of generating the fault model under multiple target scenarios based on the target scenario epoch interval and the original GNSS road test data can be implemented through the following steps:

[0144] On the basis of Figure 2 , refer to Figure 6 , Figure 6 Fig. 5 shows a flowchart of the GNSS data fault simulation method provided by the embodiment of the present application, and the step of generating the fault model under multiple target scenarios based on the target scenario epoch interval and the fault data in the target scenario epoch interval in the S240 step, which includes:

[0145] S241, determining the cycle slip data and the multipath data under the target scenario from the original GNSS road test data based on the target scenario epoch interval.

[0146] S242, performing multiple distribution fittings on the cycle slip data and the multipath data, respectively.

[0147] Among them, each distribution fitting corresponds to a distribution function.

[0148] S243, calculating the chi-square test quantity of each distribution fitting based on the cycle slip data and the distribution function corresponding to each distribution fitting, and taking the distribution function corresponding to the smallest chi-square test quantity as the cycle slip data fault model function under the target scenario, so as to generate the cycle slip data fault model under the target scenario.

[0149] S244, calculate a chi-square test quantity of each of the distribution fittings based on the multipath data and the corresponding distribution function of each of the distribution fittings, and take the distribution function corresponding to the smallest chi-square test quantity as the multipath data fault model function under the target scenario to generate the multipath data fault model under the target scenario.

[0150] S245, take the cycle slip data fault model and the multipath data fault model under the target scenario as the fault model under the target scenario.

[0151] The above steps realize the process of generating the fault model under the target scenario based on the target scenario epoch interval and the original GNSS road test data.

[0152] In the embodiment of the application, the above steps S243 and S244 can be realized by the above formula one, and the embodiment of the application will not be described again.

[0153] Optionally, a plurality of fault inclusion criteria under the target scenario are determined, and the specific process of obtaining the fault data under the target scenario according to each fault inclusion criterion through the corresponding fault model can be realized by the following steps:

[0154] On the basis of Figure 6 , please refer to Figure 7 , Figure 7 Fig. 6 shows a flowchart of a GNSS data fault simulation method provided by an embodiment of the application, and the step S250 of determining a plurality of fault inclusion criteria under the target scenario to obtain the fault data under the target scenario according to each fault inclusion criterion through the corresponding fault model, which includes:

[0155] S251, determine the cycle slip occurrence probability, the multipath occurrence fault judgment threshold and the ephemeris fault occurrence probability under the target scenario, and establish the fault inclusion criteria under the target scenario according to the cycle slip occurrence probability, the multipath occurrence fault judgment threshold and the ephemeris fault occurrence probability under the target scenario.

[0156] S254, obtain the cycle slip fault data under the target scenario according to the cycle slip occurrence probability and the cycle slip fault model function.

[0157] S255, obtain the multipath fault data under the target scenario through the multipath data fault model function under the target scenario based on the multipath occurrence fault judgment threshold under the target scenario.

[0158] S256, obtain the ephemeris fault data under the target scenario through the ephemeris fault model function under the target scenario based on the ephemeris fault occurrence probability under the target scenario.

[0159] S255, the cycle slip fault data, multipath fault data and ephemeris fault data under the target scene are taken as the fault data under the target scene.

[0160] The above steps achieve the process of determining a plurality of fault inclusion criteria under the target scene to obtain the fault data under the target scene through the corresponding fault model according to each fault inclusion criterion.

[0161] In the embodiments of the present application, the above steps can be specifically implemented by the above-mentioned formula two to formula four, and the embodiments of the present application will not be repeated here.

[0162] Optionally, the specific process of injecting the fault data under the target scene into the corresponding original GNSS road test data to obtain the GNSS fault simulation data can be implemented by the following steps:

[0163] On the basis of Figure 2 , Figure 8 , Figure 8 Fig. 7 shows a flowchart of a GNSS data fault simulation method provided by an embodiment of the present application, the feature data includes pseudorange observation value, phase observation value and ephemeris data, the step of injecting the fault data under the target scene into the corresponding original GNSS road test data to obtain the GNSS fault simulation data in the S260 step, including:

[0164] S261, injecting the fault data under the target scene into the pseudorange observation value, phase observation value and ephemeris data to obtain the GNSS fault simulation data.

[0165] Based on the same inventive concept, the embodiments of the present application also provide a GNSS data fault simulation device 300, which is used to execute the process steps in the above-mentioned embodiments and realize the corresponding technical effects.

[0166] Specifically, referring to Figure 9 , the GNSS data fault simulation device 300 is applied to an electronic device 110, and the above-mentioned GNSS data fault simulation device 300 includes an information solving module 301, a model generating module 302 and a simulation module 303.

[0167] The solving module 301 is used to obtain the original GNSS road test data under the target scene and perform RTK positioning solution to obtain a plurality of feature data.

[0168] The model generating module 302 is used to construct a decision tree classification model based on each of the feature data, identify and extract a target scene epoch interval in the original GNSS road test data through the decision tree classification model, and generate a plurality of fault models under the target scene based on the target scene epoch interval and the fault data in the target scene epoch interval.

[0169] The simulation module 303 is configured to determine a plurality of fault inclusion criteria in the target scene, to obtain fault data in the target scene according to the corresponding fault model according to each fault inclusion criterion; and to inject the fault data in the target scene into the corresponding original GNSS road test data to obtain GNSS simulation data.

[0170] Based on the same inventive concept, the embodiments of the present application also provide a computer readable storage medium, which stores a computer program, and the computer program is executed by the processor 1102 to implement the GNSS data fault simulation method provided in the above embodiments.

[0171] The steps executed by the computer program when running are not described here again, and can be referred to the above explanation and description of the GNSS data fault simulation method.

[0172] In several embodiments provided in the present application, it should be understood that the disclosed method and device can also be implemented by other manners. The above-described embodiments are only illustrative, for example, the flowchart and block diagram in the drawings show the possible implementation architecture, function and operation of the method, device and computer program product according to the embodiments of the present application. In this regard, each block in the flowchart or block diagram can represent a module, a program segment or a part of code, which contains one or more executable instructions for implementing the specified logical function. It should also be noted that in some alternative implementations, the functions annotated in the blocks can occur in different order from that annotated in the drawings. For example, two consecutive blocks can actually be executed substantially in parallel, and sometimes they can be executed in reverse order, depending on the functions involved. It should also be noted that each block in the block diagram and / or flowchart, and the combination of blocks in the block diagram and / or flowchart, can be implemented by a dedicated hardware-based system for executing the specified functions or actions, or can be implemented by a combination of dedicated hardware and computer instructions.

[0173] In addition, each functional module in the embodiments of the present application can be integrated together to form an independent part, or each module can exist independently, or two or more modules can be integrated to form an independent part.

[0174] If the functions are implemented in the form of software function modules and sold or used as independent products, they can be stored in a computer readable storage medium. Based on this understanding, the technical solutions of the present application or the parts of the technical solutions that essentially contribute to the prior art or the parts of the technical solutions can be embodied in the form of a software product. The computer software product is stored in a storage medium and includes a plurality of instructions for causing a computer device (which can be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the method described in the various embodiments of the present application. The aforementioned storage medium includes a U disk, a mobile hard disk, a read-only memory (ROM, Read-Only Memory), a random access memory (RAM, Random Access Memory), a magnetic disk or an optical disk, and various media that can store program codes.

[0175] The above only describes the preferred embodiments of the present application and is not intended to limit the present application. For those skilled in the art, the present application can have various modifications and changes. Any modification, equivalent replacement, improvement, etc. made within the spirit and principles of the present application shall be included in the protection scope of the present application.

Claims

1. A GNSS data fault simulation method, characterized in that, Be applied to electronic equipment The method comprises: Obtaining original GNSS road test data under a target scene and performing RTK positioning solution to obtain a plurality of feature data, comprising: obtaining original GNSS road test data under a target scene and performing RTK positioning solution to obtain PDOP data, pseudorange observation value number, phase observation value number, satellite number, cycle slip satellite number and average multipath data; the PDOP data, pseudorange observation value number, phase observation value number, satellite number, cycle slip satellite number and average multipath data are used as feature data; A decision tree classification model is constructed based on each feature data; The target scene epoch interval in the original GNSS road test data is identified and extracted through the decision tree classification model, comprising: obtaining the identification interval of each feature data under the target scene through the decision tree classification model, so as to identify and extract the target scene epoch interval in the original GNSS road test data according to the identification interval of each feature data; Based on the target scene epoch interval and the fault data in the target scene epoch interval, a plurality of fault models under the target scene are generated; A plurality of fault inclusion criteria under the target scene are determined, so as to obtain fault data under the target scene through the corresponding fault model according to each fault inclusion criterion; The fault data under the target scene is injected into the corresponding original GNSS road test data to obtain GNSS simulation data.

2. The GNSS data fault simulation method of claim 1, wherein, The step of constructing a decision tree classification model based on each feature data comprises: Calculate the information gain corresponding to each feature data from the root node; The feature data corresponding to the maximum information gain is used as the node division feature data; Sub-nodes are established based on the node division feature data; Recursive operation is performed on the sub-nodes to construct a decision tree classification model.

3. The GNSS data fault simulation method of claim 1, wherein, The step of generating a plurality of fault models under the target scene based on the target scene epoch interval and the fault data in the target scene epoch interval comprises: Determine the cycle slip data and multipath data under the target scene from the original GNSS road test data based on the target scene epoch interval; Various distribution fittings are performed on the cycle slip data and multipath data respectively, wherein each distribution fitting corresponds to a distribution function; Calculate the chi-square test quantity of each distribution fitting based on the cycle slip data and the distribution function corresponding to each distribution fitting, and use the distribution function corresponding to the minimum chi-square test quantity as the cycle slip data fault model function under the target scene to generate the cycle slip data fault model under the target scene; Calculate the chi-square test quantity of each distribution fitting based on the multipath data and the distribution function corresponding to each distribution fitting, and use the distribution function corresponding to the minimum chi-square test quantity as the multipath data fault model function under the target scene to generate the multipath data fault model under the target scene; The cycle slip data fault model and multipath data fault model under the target scene are used as the fault model under the target scene.

4. The GNSS data fault simulation method of claim 3, wherein, The fault data further includes ephemeris fault data, and the fault model under the target scenario further includes an ephemeris fault model; and the step of determining the plurality of fault inclusion criteria under the target scenario to obtain the fault data under the target scenario according to the corresponding fault model includes: determining a cycle slip occurrence probability, a multipath occurrence fault judgment threshold and an ephemeris fault occurrence probability under the target scenario, and establishing the fault inclusion criteria under the target scenario according to the cycle slip occurrence probability, the multipath occurrence fault judgment threshold and the ephemeris fault occurrence probability under the target scenario; obtaining cycle slip fault data under the target scenario according to the cycle slip occurrence probability and a cycle slip fault model function; obtaining multipath fault data under the target scenario according to the multipath occurrence fault judgment threshold under the target scenario and a multipath data fault model function under the target scenario; obtaining ephemeris fault data under the target scenario according to the ephemeris fault occurrence probability under the target scenario and an ephemeris fault model function under the target scenario; taking the cycle slip fault data, the multipath fault data and the ephemeris fault data under the target scenario as the fault data under the target scenario.

5. The GNSS data fault emulation method of claim 1, wherein, The original GNSS road test data includes pseudorange observation values, phase observation values and ephemeris data; The step of injecting the fault data under the target scenario into the corresponding original GNSS road test data to obtain GNSS fault simulation data includes: injecting the fault data under the target scenario into the pseudorange observation values, the phase observation values and the ephemeris data to obtain GNSS fault simulation data.

6. A GNSS data fault simulation apparatus characterized by comprising: Applied to an electronic device; The device includes: a solving module configured to obtain original GNSS road test data under a target scenario and perform RTK positioning solving to obtain a plurality of feature data, including: obtaining original GNSS road test data under a target scenario and performing RTK positioning solving to obtain PDOP data, a number of pseudorange observation values, a number of phase observation values, a number of satellites, a number of cycle slip satellites and average multipath data; and taking the PDOP data, the number of pseudorange observation values, the number of phase observation values, the number of satellites, the number of cycle slip satellites and the average multipath data as feature data; a model generating module configured to construct a decision tree classification model based on the feature data, and identify and extract a target scenario epoch interval in the original GNSS road test data through the decision tree classification model, including: obtaining an identification interval of each feature data under the target scenario through the decision tree classification model, and identifying and extracting a target scenario epoch interval in the original GNSS road test data according to the identification interval of each feature data; and generating a plurality of fault models under the target scenario based on the target scenario epoch interval and fault data in the target scenario epoch interval. The simulation module is configured to determine a plurality of fault inclusion criteria in the target scenario, to obtain fault data in the target scenario according to the corresponding fault model according to each fault inclusion criterion; and to inject the fault data in the target scenario into the corresponding original GNSS road test data to obtain GNSS simulation data.

7. An electronic device, comprising: comprising a memory and a processor; the memory is configured to store a computer program; the processor is configured to execute the computer program to implement the GNSS data fault simulation method according to any one of claims 1-5.

8. A computer-readable storage medium having stored thereon a computer program, characterized in that, The computer program, when executed by the processor, implements the GNSS data fault simulation method according to any one of claims 1-5.

Images