CN-122015819-A - Real-time vehicle renting positioning method and system based on quantum positioning

CN122015819ACN 122015819 ACN122015819 ACN 122015819ACN-122015819-A

Abstract

The invention relates to the technical field of inertial navigation and multisource sensor fusion, and discloses a real-time positioning method and system for a rented vehicle based on quantum positioning, wherein a vehicle-mounted computing unit synchronously acquires first angular rate data of a vehicle-mounted micro-electromechanical inertial measurement unit and second angular rate data of a vehicle-mounted chip-level atomic spin gyroscope; the method comprises the steps of utilizing first angular rate data to execute strapdown inertial calculation, establishing a sliding integral window to execute numerical accumulation and downsampling operation on the first angular rate data to generate an equivalent reconstruction angular rate which is physically isomorphic with second angular rate data, calculating integral matching residual errors, constructing a measurement equation, inputting error state Kalman filter to calculate and feed back zero offset estimated values.

Inventors

SUN PENG

Assignees

北京嘉朋天地汽车服务有限公司

Dates

Publication Date: 20260512
Application Date: 20251226

Claims (10)

1. A real-time vehicle renting positioning method based on quantum positioning is characterized by comprising the following steps: The method comprises the steps of data acquisition, wherein a vehicle-mounted computing unit synchronously acquires first angular rate data output by a vehicle-mounted micro-electromechanical inertial measurement unit and second angular rate data output by a vehicle-mounted chip-level atomic spin gyroscope, the sampling frequency of the first angular rate data is higher than that of the second angular rate data, and the second angular rate data represents the average angular rate of an atomic ensemble in a sampling period; a strapdown calculating step, namely performing strapdown inertial calculation by using the first angular rate data, updating the attitude quaternion, speed and position of the vehicle, and generating a priori navigation state vector; A time domain integration reconstruction step of establishing a sliding integration window aligned with a sampling period of the second angular rate data in a time dimension, acquiring a plurality of first angular rate data points contained in the sliding integration window, performing numerical accumulation and downsampling operation, and generating an equivalent reconstruction angular rate isomorphic with the second angular rate data in a physical sense; A residual error construction and observation step, namely calculating a difference value between the equivalent reconstruction angular rate and the second angular rate data to obtain an integral matching residual error for eliminating a phase lag error of the dynamic response, and constructing a measurement equation about the zero offset error of the micro-electromechanical inertial measurement unit based on the integral matching residual error; And a closed loop correction step, namely inputting the measurement equation into a preset error state Kalman filter, calculating a zero offset estimated value of the micro-electromechanical inertial measurement unit, performing depolarization processing on the first angular rate data by using the zero offset estimated value, correcting the prior navigation state vector, and simultaneously storing the zero offset estimated value as a feedback input variable at the next moment.
2. The real-time positioning method of a vehicle renting vehicle based on quantum positioning according to claim 1, wherein in the time domain integral reconstruction step, the equivalent reconstruction angular rate is calculated based on the following formula: , wherein, In order to reconstruct the angular rate equally, Is a frequency multiple of the first angular rate data relative to the second angular rate data, To the first in the sliding integration window A first one of the data points of the angular rate, And outputting and storing the zero offset estimated value for the Kalman filter of the error state at the last moment.
3. The real-time positioning method for a vehicle renting based on quantum positioning according to claim 1, further comprising an initial course alignment step before the strapdown calculation step, wherein the static base judgment step comprises the steps of monitoring a specific force variance and an angular rate variance output by a micro-electromechanical inertial measurement unit, judging that the vehicle is in a static base state when the specific force variance and the angular rate variance are lower than preset static threshold values, the earth rotation vector extraction step comprises the steps of obtaining a time average angular rate vector output by a chip-level atomic spin gyroscope in the static base state, identifying the time average angular rate vector as a projection component of the earth rotation angular rate in a vehicle carrier coordinate system based on a preset earth rotation angular rate module value, and the absolute course calculation step comprises the steps of determining a horizontal attitude plane by combining a gravity vector measured by the micro-electromechanical inertial measurement unit, calculating an included angle between the direction of the projection component in the horizontal attitude plane and a vehicle longitudinal axis, obtaining an absolute true north course angle of the vehicle, and injecting a priori navigation state vector.
4. The real-time positioning method for the rented vehicle based on quantum positioning according to claim 1 is characterized by further comprising a low-dynamic non-integrity constraint step of working condition identification, wherein the working condition identification comprises the steps of monitoring the specific force fluctuation characteristic of a wheel speed scalar and a micro-electromechanical inertia measurement unit of the vehicle in real time, activating a virtual observation mode when the wheel speed scalar is in a non-zero low-speed interval and the specific force fluctuation characteristic is lower than a preset weak observation threshold value, constructing a virtual lateral speed observation value and a virtual vertical speed observation value with zero values based on a non-integrity constraint principle of vehicle kinematics, and constraint injection, wherein the speed residuals of the instantaneous lateral speed and the instantaneous vertical speed derived from a priori navigation state vector relative to the virtual lateral speed observation value and the virtual vertical speed observation value are calculated, and the speed residuals are used as auxiliary observation values to be input into an error state Kalman filter.
5. The real-time positioning method for a vehicle renting based on quantum positioning according to claim 1 is characterized in that in the closed loop correction step, the state vector of an error state Kalman filter at least comprises an attitude error, a speed error, a position error, a gyroscope zero bias error and an accelerometer zero bias error, the closed loop correction step further comprises covariance prediction, namely predicting an error covariance matrix of a state vector by utilizing noise parameters of first angular rate data and a first-order Markov process model, gain calculation and updating, namely calculating Kalman gain by combining the noise matrix of a measurement equation, and mapping an integral matching residual error into a state vector space by utilizing the Kalman gain to obtain an optimal estimated value of each error state, wherein the optimal estimated value of the gyroscope zero bias error is the zero bias estimated value.
6. The real-time positioning method for the rented vehicle based on quantum positioning according to claim 1, wherein the data acquisition step further comprises a time stamp hard alignment process, wherein a unified trigger pulse signal is generated by a hardware timer of the vehicle-mounted computing unit, and the data latching operation of the micro-electromechanical inertial measurement unit and the chip-level atomic spin gyroscope is triggered simultaneously by the trigger pulse signal, so that the first angular rate data and the second angular rate data are ensured to have the same time reference.
7. The real-time positioning method for the rented vehicle based on quantum positioning according to claim 1 is characterized by further comprising a residual error checking step of calculating a normalized chi-square value of an integral matching residual error, comparing the normalized chi-square value with a preset fault detection threshold value, judging that measured data are abnormal when the normalized chi-square value exceeds the fault detection threshold value, and blocking the input of the integral matching residual error to an error state kalman filter at the present time, and maintaining positioning output only by using a priori navigation state vector.
8. The real-time positioning method for the vehicle renting based on quantum positioning according to claim 1 is characterized in that an error state feedback correction mode is adopted in the closed loop correction step, an error state optimal estimated value output by an error state Kalman filter is fed back to a navigation state quantity accumulated in the strapdown calculation step when a filtering period is finished, an error state vector in the error state Kalman filter is cleared after feedback correction is finished, an error covariance matrix is reserved for recursion of the next period, a chip-level atomic spin gyroscope only senses the angular rate of a yaw axis direction of the vehicle, a measurement equation is constructed only for errors of the yaw angular rate dimension in the residual construction and observation step, and the error state Kalman filter estimates and corrects zero bias errors of a horizontal axial micro-electromechanical gyroscope through a coupling relation of the state transition matrix by using observation information of the yaw angular rate dimension.
9. The real-time positioning method for a vehicle renting based on quantum positioning according to claim 1, further comprising a zero-speed static resetting step of suspending position updating of the strapdown resolving step when the wheel speed of the vehicle is monitored to be zero and the duration exceeds a preset resetting window, calibrating zero bias of the micro-electromechanical inertial measurement unit by using an output average value of the chip-level atomic spin gyroscope in a static state, and resetting a zero bias state initial value in an error state Kalman filter by using a calibration result.
10. A real-time positioning system for a rented vehicle based on quantum positioning for implementing the method of claim 1, characterized in that the system comprises an on-board microelectromechanical inertial measurement unit, an on-board chip atomic spin gyroscope, and an on-board computing unit: The vehicle-mounted computing unit comprises a data acquisition module, a strapdown resolving module, a time domain integral reconstruction module, a residual error construction and observation module and a closed loop correction module; The data acquisition module is used for synchronously acquiring first angular rate data output by the vehicle-mounted micro-electromechanical inertial measurement unit and second angular rate data output by the vehicle-mounted chip-level atomic spin gyroscope, wherein the sampling frequency of the first angular rate data is higher than that of the second angular rate data, and the second angular rate data represents the average angular rate of the atomic ensemble in a sampling period; The strapdown resolving module is used for executing strapdown inertial resolving by utilizing the first angular rate data, updating the attitude quaternion, speed and position of the vehicle and generating a priori navigation state vector; The time domain integral reconstruction module is used for establishing a sliding integral window aligned with the sampling period of the second angular rate data in the time dimension, acquiring a plurality of first angular rate data points contained in the sliding integral window, performing numerical accumulation and downsampling operation, and generating an equivalent reconstruction angular rate isomorphic with the second angular rate data in the physical meaning; The residual construction and observation module is used for calculating the difference between the equivalent reconstruction angular rate and the second angular rate data, obtaining an integral matching residual for eliminating the phase lag error of the dynamic response, and constructing a measurement equation about the zero offset error of the micro-electromechanical inertial measurement unit based on the integral matching residual; the closed loop correction module is used for inputting the measurement equation into a preset error state Kalman filter, calculating a zero offset estimated value of the micro-electromechanical inertial measurement unit, performing depolarization processing on the first angular rate data by using the zero offset estimated value, and correcting the prior navigation state vector.

Description

Real-time vehicle renting positioning method and system based on quantum positioning Technical Field The invention relates to a real-time positioning method and a system for a rented vehicle based on quantum positioning, and belongs to the technical field of inertial navigation and multi-source sensor fusion. Background The continuous high-precision positioning requirements of the current time-sharing leasing vehicles in satellite signal rejection environments of underground garages, multilayer three-dimensional parking buildings and the like are urgent, the main stream scheme of the industry adopts a global satellite navigation system to cooperate with a micro-electromechanical inertial measurement unit to maintain track calculation during vehicle signal occlusion, the physical characteristics of limited materials of a silicon-based micro-electromechanical sensor exist, random walk noise and zero bias instability exist, positioning errors are non-linearly dispersed along with time under the working condition of no external absolute position reference for a long period of time, and billing disputes and asset monitoring failure are caused. The atomic spin effect chip-level quantum sensor is used as an error observation reference to inhibit inertial drift, the atomic sensor has excellent zero bias stability, and besides the limitation of sensor hardware performance, the related control method also has the defects, for example, the Chinese patent publication No. CN110631575B discloses a strapdown system based on an atomic spin gyroscope, the scheme adopts the combination of a rate gyroscope and two-degree-of-freedom atomic spin gyroscopes, the orthogonal coupling error of the atomic spin gyroscope is compensated by using the measured value of the rate gyroscope, the method improves the measurement precision under quasi-static state, but does not solve the phase lag problem caused by the response bandwidth mismatch of the heterogeneous sensor, because of the lack of an active compensation mechanism for physical bandwidth difference, when a vehicle turns continuously or turns spirally, a system is difficult to distinguish dynamic lag errors from inherent zero offset of devices, so that navigation solution divergence is limited by spin polarization in an atomic gas chamber and physical relaxation effect in a detection process, dynamic response bandwidth is lower than that of a micro-electromechanical sensor, when a leased vehicle executes out-of-warehouse spiral climbing and continuous turning high-dynamic maneuver, physical phase lag is generated by frequency domain response mismatch among heterogeneous sensors, a conventional instantaneous differential data fusion strategy is adopted, a filtering algorithm is difficult to distinguish dynamic lag errors from inherent zero offset of devices, and false correction quantity is introduced to destroy navigation solution convergence. Therefore, how to overcome the drift divergence in the dynamic response mismatch and weak observation environment by a signal processing mechanism based on a low-cost heterogeneous hardware architecture and realize full-working-condition high-precision autonomous navigation becomes the technical problem to be solved by the invention. Disclosure of Invention In order to solve the problems in the background technology, the technical scheme of the invention is as follows, a real-time positioning method for a rented vehicle based on quantum positioning, which comprises the following steps: The method comprises the steps of data acquisition, wherein a vehicle-mounted computing unit synchronously acquires first angular rate data output by a vehicle-mounted micro-electromechanical inertial measurement unit and second angular rate data output by a vehicle-mounted chip-level atomic spin gyroscope, the sampling frequency of the first angular rate data is higher than that of the second angular rate data, and the second angular rate data represents the average angular rate of an atomic ensemble in a sampling period; a strapdown calculating step, namely performing strapdown inertial calculation by using the first angular rate data, updating the attitude quaternion, speed and position of the vehicle, and generating a priori navigation state vector; A time domain integration reconstruction step of establishing a sliding integration window aligned with a sampling period of the second angular rate data in a time dimension, acquiring a plurality of first angular rate data points contained in the sliding integration window, performing numerical accumulation and downsampling operation, and generating an equivalent reconstruction angular rate isomorphic with the second angular rate data in a physical sense; A residual error construction and observation step, namely calculating a difference value between the equivalent reconstruction angular rate and the second angular rate data to obtain an integral matching residual error for eliminating a phase lag error of