技术领域technical field
本发明涉及一种基于模型摄动的火星大气进入自适应估计方法，属于深空探测技术领域。The invention relates to an adaptive estimation method for the entry of Martian atmosphere based on model perturbation, and belongs to the technical field of deep space exploration.
背景技术Background technique
未来火星采样返回及载人探测任务要求探测器具备表面定点着陆的能力，进入段主动制导与控制是实现火星定点着陆的有效途径。而探测器制导系统需要导航系统为其提供精确的状态信息以保证下传给控制系统指令的精度,这就要求进入过程中探测器能够实现高精度自主导航。The future Mars sampling return and manned exploration missions require the probe to have the ability to land at a fixed point on the surface. Active guidance and control in the entry segment is an effective way to achieve fixed-point landing on Mars. The detector guidance system needs the navigation system to provide it with accurate status information to ensure the accuracy of the instructions transmitted to the control system, which requires the detector to achieve high-precision autonomous navigation during the entry process.
目前火星大气进入段采用基于惯性测量单元(Inertial Measurement Unit,IMU)输出的航位递推导航方法，但由于该方法测量信息单一，不能对初始状态偏差进行修正，且受敏感器自身系统噪声的影响，不能满足未来火星定点着陆对导航系统的精度要求。针对航位递推导航方法存在的不足，学者相继提出了引入不同外部测量信息结合IMU输出构建火星大气进入段组合导航方法，来修正进入点初始偏差。但新的导航信息的引入也带来了新的问题，在引入外部测量信息的同时也将进入动力学模型引入了状态估计过程中。当动力学系统模型参数与实际飞行过程中真实模型参数存在较大偏差时，不确定参数引起的摄动会降低导航系统的性能，甚至会导致状态估计误差的发散。在火星大气进入组合导航方法中，对动力学模型产生摄动的主要因素有火星大气密度及探测器气动力系数。因此，如何有效抑制组合导航方法中模型摄动对导航性能的影响是提高进入段状态估计精度面临的主要问题。At present, the dead-recursive navigation method based on the output of the Inertial Measurement Unit (IMU) is used in the entry stage of the Martian atmosphere. However, due to the single measurement information of this method, the initial state deviation cannot be corrected, and it is affected by the noise of the sensor's own system. It cannot meet the precision requirements of the future Mars fixed-point landing on the navigation system. In view of the shortcomings of the dead recursive navigation method, scholars have successively proposed the introduction of different external measurement information combined with the output of the IMU to construct an integrated navigation method for the entry segment of the Martian atmosphere to correct the initial deviation of the entry point. However, the introduction of new navigation information also brings new problems. When the external measurement information is introduced, the dynamic model is also introduced into the state estimation process. When there is a large deviation between the model parameters of the dynamic system and the real model parameters in the actual flight process, the perturbation caused by the uncertain parameters will reduce the performance of the navigation system, and even lead to the divergence of the state estimation error. In the integrated navigation method of the Martian atmosphere entering, the main factors that perturb the dynamic model are the density of the Martian atmosphere and the aerodynamic coefficient of the probe. Therefore, how to effectively suppress the influence of model perturbation on the navigation performance in the integrated navigation method is the main problem to improve the state estimation accuracy of the entry segment.
发明内容SUMMARY OF THE INVENTION
本发明公开的一种基于模型摄动的火星大气进入自适应估计方法，要解决的技术问题是减小火星进入段组合导航方法中动力学模型摄动对状态估计精度的影响，保证进入过程中探测器状态估计精度及导航系统的稳定性。The invention discloses a model perturbation-based adaptive estimation method for the entry of the Martian atmosphere, and the technical problem to be solved is to reduce the influence of the dynamic model perturbation on the state estimation accuracy in the integrated navigation method of the Mars entry segment, and to ensure that during the entry process Detector state estimation accuracy and navigation system stability.
本发明的目的是通过下述技术方案实现的。The purpose of the present invention is achieved through the following technical solutions.
本发明公开的一种基于模型摄动的火星大气进入自适应估计方法，首先通过分析火星大气进入段不确定参数与气动力模型的耦合关系，把不确定参数对动力学系统的摄动转化为气动力模型偏差，减少动力学系统中不确定参数项个数，所述的火星大气进入段不确定参数包括火星大气密度与气动力系数。然后，根据动力学系统中气动力模型摄动的变化规律，建立针对动力学系统中摄动量偏差的滤波模型，所述的模型摄动量变化范围为有界的，每个滤波模型动力学系统中摄动量对应于有界范围内的一个偏差值，根据每个滤波模型预测测量残差，并基于测量残差信息自适应更新各模型的权值，不断迭代逼近真实模型摄动，从而抑制气动力模型偏差对状态估计精度的影响，保证进入过程中探测器状态估计精度及导航系统的稳定性。The invention discloses an adaptive estimation method for Mars atmosphere entry based on model perturbation. First, by analyzing the coupling relationship between the uncertain parameters and the aerodynamic model of the Mars atmosphere entry section, the perturbation of the uncertain parameters to the dynamic system is transformed into The deviation of the aerodynamic model reduces the number of uncertain parameter items in the dynamic system. The uncertain parameters in the entry section of the Martian atmosphere include the density of the Martian atmosphere and the aerodynamic coefficient. Then, according to the change rule of the perturbation of the aerodynamic model in the dynamic system, a filter model for the deviation of the perturbation amount in the dynamic system is established, and the variation range of the model perturbation amount is bounded. The amount of perturbation corresponds to a deviation value within a bounded range. According to each filter model, the measurement residual is predicted, and the weights of each model are adaptively updated based on the measurement residual information, and the perturbation of the real model is continuously iteratively approached, thereby suppressing the aerodynamic force. The influence of the model deviation on the state estimation accuracy ensures the state estimation accuracy of the detector and the stability of the navigation system during the entry process.
所述的一种基于模型摄动的火星大气进入自适应估计方法，能够应用于火星大气进入段动压测量辅助的组合导航方法或无线电/IMU组合导航方法等，提高导航状态估计的精度和系统的稳定性。The described method for self-adaptive estimation of the entry of the Martian atmosphere based on model perturbation can be applied to the combined navigation method assisted by the dynamic pressure measurement of the Mars atmosphere entering segment or the radio/IMU combined navigation method, etc., so as to improve the accuracy and system of the estimation of the navigation state. stability.
本发明公开的一种基于模型摄动的火星大气进入自适应估计方法，包括如下步骤：A method for self-adaptive estimation of Martian atmosphere entry based on model perturbation disclosed in the present invention includes the following steps:
步骤一、建立基于火星惯性系的动力学模型。Step 1. Establish a dynamic model based on the Mars inertial system.
为了简化模型，假设进入过程中探测器在配平功角条件下飞行，侧滑角为零，且控制量倾侧角为零；进入段基于火星惯性系的动力学模型建立如公式(1)In order to simplify the model, it is assumed that the probe flies under the condition of the trim power angle during the entry process, the sideslip angle is zero, and the control variable inclination angle is zero; the dynamic model of the entry segment based on the Mars inertial system is established as formula (1)
其中，in,
公式(2)、(3)中，L/D表示探测器升阻比，B＝m/CDS表示弹道系数，CD表示探测器的阻力系数，S表示探测器参考面积，m表示探测器质量。探测器升阻比L/D、弹道系数B，阻力系数CD，参考面积S以及质量m的标称值都视为已知。ρ为火星大气密度，假设火星大气密度ρ呈指数形式分布形式且满足公式(4)In formulas (2) and (3), L/D represents the lift-to-drag ratio of the detector, B=m/CD S represents the ballistic coefficient,CD represents the drag coefficient of the detector, S represents the reference area of the detector, and m represents the detection device quality. The lift-to-drag ratio L/D of the detector, the ballistic coefficient B, the drag coefficient CD , the reference area S and the nominal value of the mass m are all considered to be known. ρ is the density of the Martian atmosphere, assuming that the density ρ of the Martian atmosphere is exponentially distributed and satisfies the formula (4)
其中，ρ0＝2e-4kg/m3表示火星大气参考密度，r0＝3,437,200m表示参考高度，hs＝7500m火星大气标高。Among them, ρ0 =2e-4kg/m3 represents the reference density of the Martian atmosphere, r0 =3,437,200m represents the reference altitude, and hs =7500m Mars atmospheric elevation.
步骤二、建立进入段摄动模型。Step 2: Establish an entry-segment perturbation model.
由进入段动力学系统可知，火星大气密度ρ及探测器气动力系数与动力学系统中气动力模型紧密相关。而在组合导航方法状态估计中，不确定参数引起的摄动会随着动力学方程递推传播到下一时刻的预估状态及误差协方差中，从而导致量测更新增益出现偏差，不能得到最优状态估计。而在火星大气进入段，大气密度ρ及弹道系数B随时间存在较大的不确定性，由公式(2)、(3)知，火星大气密度ρ及气动力系数耦合在气动力模型中，为了处理不确定参数引起的模型摄动对导航状态估计的影响，把大气密度ρ和弹道系数B对动力学的摄动归结于参数τ定义如下，From the dynamic system of the entry stage, it can be known that the Martian atmospheric density ρ and the aerodynamic coefficient of the probe are closely related to the aerodynamic model in the dynamic system. However, in the state estimation of the integrated navigation method, the perturbation caused by the uncertain parameters will be recursively propagated to the estimated state and error covariance at the next moment along with the dynamic equation, resulting in the deviation of the measurement update gain, which cannot be obtained. optimal state estimation. However, in the entry stage of the Martian atmosphere, the atmospheric density ρ and the ballistic coefficient B have great uncertainty with time. From formulas (2) and (3), it is known that the Martian atmospheric density ρ and the aerodynamic coefficient are coupled in the aerodynamic model, In order to deal with the influence of the model perturbation caused by the uncertain parameters on the estimation of the navigation state, the perturbation of the atmospheric density ρ and the ballistic coefficient B on the dynamics is attributed to the parameter τ, which is defined as follows:
大气密度ρ和弹道系统不确定性对参数τ的影响可由式(6)表示，The influence of atmospheric density ρ and ballistic system uncertainty on parameter τ can be expressed by Eq. (6),
公式(6)中，右上角带*的变量表示机载模型参数，为定值；Δ表示机载模型与真实模型的偏差，在实际飞行过程中是未知时变的，但变化范围假设为已知的。因此，大气密度ρ及弹道系数B不确定性对动力学系统的摄动在气动力模型中的体现公式(7)和公式(8)。In formula (6), the variable with * in the upper right corner represents the airborne model parameter, which is a fixed value; Δ represents the deviation between the airborne model and the real model, which is unknown and time-varying in the actual flight process, but the variation range is assumed to be Known. Therefore, the perturbation of the atmospheric density ρ and the uncertainty of the ballistic coefficient B to the dynamic system is reflected in the aerodynamic model by formula (7) and formula (8).
由公式(7)和公式(8)可知，把不确定参数大气密度ρ及弹道系数B引起的摄动转换为动力学系统中气动力的模型偏差，能够减少动力学系统中不确定项的个数，且气动力偏差的变化规律与不确定参数变化规律相同。From formula (7) and formula (8), it can be known that converting the perturbation caused by the uncertain parameters atmospheric density ρ and ballistic coefficient B into the model deviation of the aerodynamic force in the dynamic system can reduce the number of uncertain items in the dynamic system. and the change law of the aerodynamic deviation is the same as the change law of the uncertain parameters.
步骤三、利用自适应估计方法抑制不确定参数对动力学系统的扰动，提高位置和速度估计精度。Step 3: Use the adaptive estimation method to suppress the disturbance of the uncertain parameters to the dynamic system, and improve the position and velocity estimation accuracy.
根据步骤二中建立的模型摄动的有界性及其变化规律，建立针对动力学系统中摄动量Δτ的滤波模型，所述的摄动量Δτ变化范围固定且有界，每个滤波模型动力学系统中摄动量Δτ对应于有界范围内的一个偏差值，根据每个滤波模型预测测量残差，并基于测量残差信息自适应更新各模型的权值，最大权值对应的滤波模型表示该模型中摄动值最接近真实的动力学系统的摄动值。并对火星大气进入段测量信息进行无量纲化处理，无量纲化处理后的自适应估计方法各模型权值的求取方法如公式(9)所示，According to the boundedness of the model perturbation established in step 2 and its variation rule, a filter model for the perturbationΔτ in the dynamic system is established. The variation range of the perturbationΔτ is fixed and bounded. Each filter model In the dynamic system, the perturbationΔτ corresponds to a deviation value within a bounded range. The measurement residual is predicted according to each filtering model, and the weights of each model are adaptively updated based on the measurement residual information. The filter corresponding to the maximum weight The model represents the perturbation value in the model that is closest to the real dynamic system. The measurement information of the entry section of the Martian atmosphere is subjected to dimensionless processing. The method for calculating the weights of each model of the adaptive estimation method after the dimensionless processing is shown in formula (9).
权值满足公式(10)所述的条件，weight Satisfies the condition described in formula (10),
式中，表示k时刻的无量纲化后的测量信息，Z0为已知常量，ai表示第i个滤波模型测量信息对应的输入权重。标量ui表示改进后第i个滤波模型与当前测量信息的匹配度。In the formula, Represents the dimensionless measurement information at time k, Z0 is a known constant, and ai represents the input weight corresponding to the measurement information of the ith filter model. The scalarui represents the matching degree between the ith filtering model after improvement and the current measurement information.
无量纲化处理后各模型测量信息权重更新方法如公式(11)所示，After dimensionless processing, the update method of the measurement information weight of each model is shown in formula (11),
式中，η表示学习律，根据实际情况由使用者自定义，hi为，In the formula, η represents the learning law, which is defined by the user according to the actual situation, andhi is,
第i个滤波模型的后验概率密度函数，且有The posterior probability density function of the ith filter model, and has
式中，In the formula,
通过公式(9)求取各滤波模型权值、并通过公式(11)更新各滤波模型的权值能够自适应逼近真实模型摄动，精确确定摄动量在每个采样时刻的摄动值，从而抑制摄动对位置和速度估计的影响，提高导航状态估计的精度和系统的稳定性。Obtaining the weights of each filter model by formula (9) and updating the weights of each filter model by formula (11) can adaptively approximate the perturbation of the real model, and accurately determine the perturbation value of the perturbation amount at each sampling moment, so that The influence of perturbation on position and velocity estimation is suppressed, and the accuracy of navigation state estimation and the stability of the system are improved.
为提高解算速度，满足进入段导航实时性要求，所述的滤波模型优选采用EKF对探测器的位置r和速度v进行解算。In order to improve the calculation speed and meet the real-time requirement of navigation in the entry segment, the filtering model preferably adopts EKF to calculate the position r and velocity v of the detector.
所述的一种基于模型摄动的火星大气进入自适应估计方法，能够应用于火星大气进入段动压测量辅助的组合导航方法或无线电/IMU组合导航方法等，提高导航状态估计的精度和系统的稳定性。The described method for self-adaptive estimation of the entry of the Martian atmosphere based on model perturbation can be applied to the combined navigation method assisted by the dynamic pressure measurement of the Mars atmosphere entering segment or the radio/IMU combined navigation method, etc., so as to improve the precision and system of the estimation of the navigation state. stability.
传统的模型权值求取方法直接采用对测量信息进行不同加权来求取各滤波应模型的权值。但在火星大气进入段，峰值动压、加速度及探测器与无线电信标之间的相对距离及速度等观测量数值都很大，采用传统的权重求取方法会使部分模型权重趋于无穷大，导致数值计算问题。步骤三对火星进入段测量信息进行无量纲化处理，能够保证权值求取的数值稳定性。The traditional method for calculating the model weights directly uses different weightings on the measurement information to obtain the weights of each filtering model. However, in the entry stage of the Martian atmosphere, the peak dynamic pressure, acceleration, and the relative distance and velocity between the probe and the radio beacon are all very large. The traditional weight calculation method will make the weight of some models tend to infinity. cause numerical problems. In step 3, dimensionless processing is performed on the measurement information of the Mars entry segment, which can ensure the numerical stability of the weight calculation.
有益效果：Beneficial effects:
1、本发明公开的一种基于模型摄动的火星大气进入自适应估计方法，通过把不确定参数(大气密度ρ和弹道系数B)对动力学系统的摄动转化为气动力模型偏差，减少动力学系统中不确定参数项个数。1. A model-based perturbation-based approach to self-adaptive estimation of Martian atmospheric entry disclosed in the present invention reduces the perturbation of the dynamic system by uncertain parameters (atmospheric density ρ and ballistic coefficient B) into aerodynamic model deviations. The number of uncertain parameter terms in a dynamical system.
2、本发明公开的一种基于模型摄动的火星大气进入自适应估计方法，通过对测量信息进行无量纲化处理，能够自适应求取每个滤波模型的权值，不断迭代逼近真实的模型摄动,从而抑制不确定参数(大气密度ρ和弹道系数B)引起的动力学扰动对状态估计精度的影响，保证进入过程中探测器状态估计精度及导航系统的稳定性。2. A method for self-adaptive estimation of Martian atmosphere entry based on model perturbation disclosed in the present invention, by performing dimensionless processing on measurement information, the weight value of each filter model can be adaptively obtained, and the real model can be continuously iteratively approached. Therefore, the influence of dynamic disturbance caused by uncertain parameters (atmospheric density ρ and ballistic coefficient B) on the state estimation accuracy is suppressed, and the state estimation accuracy of the detector and the stability of the navigation system are ensured during the entry process.
3、本发明公开的一种基于模型摄动的火星大气进入自适应估计方法，采用EKF对滤波模型位置r和速度v进行解算，能够提高解算速度，满足进入段导航实时性要求。3. The present invention discloses a method for self-adaptive estimation of Martian atmospheric entry based on model perturbation. The EKF is used to calculate the position r and velocity v of the filter model, which can improve the calculation speed and meet the real-time requirements of navigation in the entry segment.
附图说明Description of drawings
图1为压力传感器分布示意图。图1(a)表示压力传感器在探测器上的位置，图1(b)表示传感器轴线与进入速度方向的空间关系。Figure 1 is a schematic diagram of the distribution of pressure sensors. Figure 1(a) shows the position of the pressure sensor on the detector, and Figure 1(b) shows the spatial relationship between the sensor axis and the direction of the incoming velocity.
图2为火星大气进入自主导航策略流程图。Figure 2 is a flow chart of the autonomous navigation strategy for entering the Martian atmosphere.
图3为本发明提出的自适应估计方法状态估计性能图。图3(a)表示x轴位置估计误差及3σ偏差，图3(b)表示y轴位置估计误差及3σ偏差，图3(c)表示z轴位置估计误差及3σ偏差，图3(d)表示x轴速度估计误差及3σ偏差，图3(e)表示y轴速度估计误差及3σ偏差，图3(f)表示z轴速度估计误差及3σ偏差。FIG. 3 is a state estimation performance diagram of the adaptive estimation method proposed by the present invention. Figure 3(a) shows the x-axis position estimation error and 3σ deviation, Figure 3(b) shows the y-axis position estimation error and 3σ deviation, Figure 3(c) shows the z-axis position estimation error and 3σ deviation, and Figure 3(d) Indicates the x-axis velocity estimation error and 3σ deviation, Fig. 3(e) shows the y-axis velocity estimation error and 3σ deviation, and Fig. 3(f) shows the z-axis velocity estimation error and 3σ deviation.
具体实施方式Detailed ways
为了更好的说明本发明的目的和优点，下面结合附图和实例对发明内容做进一步说明。In order to better illustrate the purpose and advantages of the present invention, the content of the invention will be further described below with reference to the accompanying drawings and examples.
实施例1：采用火星进入段基于动压测量辅助的组合导航方法为实例分析，该导航方法动力学模型及量测模型参数都具有较大不确定性，更能体现本实施例的实用性。本实施例公开的一种基于模型摄动的火星大气进入自适应估计方法，具体实施方法包括如下步骤：Embodiment 1: The combined navigation method based on dynamic pressure measurement assistance in the Mars entry segment is used as an example for analysis. The dynamic model and measurement model parameters of the navigation method have large uncertainties, which can better reflect the practicability of this embodiment. A method for self-adaptive estimation of the entry of the Martian atmosphere based on model perturbation disclosed in this embodiment, the specific implementation method includes the following steps:
为了简化模型，假设进入过程中探测器在配平功角条件下飞行，侧滑角为零，且控制量倾侧角为零；进入段基于火星惯性系的动力学模型建立如下，In order to simplify the model, it is assumed that the probe flies under the condition of the trim power angle during the entry process, the sideslip angle is zero, and the control variable inclination angle is zero; the dynamic model of the entry segment based on the Mars inertial system is established as follows:
其中，in,
公式(2)、(3)中，L/D表示探测器升阻比，B＝m/CDS表示弹道系数，CD表示探测器的阻力系数，S表示探测器参考面积，m表示探测器质量。探测器升阻比L/D、弹道系数B，阻力系数CD，参考面积S以及质量m的标称值都视为已知。ρ为火星大气密度，假设火星大气密度ρ呈指数形式分布形式且满足公式(4)In formulas (2) and (3), L/D represents the lift-to-drag ratio of the detector, B=m/CD S represents the ballistic coefficient,CD represents the drag coefficient of the detector, S represents the reference area of the detector, and m represents the detection device quality. The lift-to-drag ratio L/D of the detector, the ballistic coefficient B, the drag coefficient CD , the reference area S and the nominal value of the mass m are all considered to be known. ρ is the density of the Martian atmosphere, assuming that the density ρ of the Martian atmosphere is exponentially distributed and satisfies the formula (4)
其中，ρ0＝2e-4kg/m3表示火星大气参考密度，r0＝3,437,200m表示参考高度，hs＝7500m火星大气标高。Among them, ρ0 =2e-4kg/m3 represents the reference density of the Martian atmosphere, r0 =3,437,200m represents the reference altitude, and hs =7500m Mars atmospheric elevation.
在火星进入过程中，IMU可以实时测量作用于探测器上的气动加速度，本实施例忽略了加速度计的尺度因子偏差及非校准偏差，加速度计测量模型如公式(5)所示，During the Mars entry process, the IMU can measure the aerodynamic acceleration acting on the probe in real time. In this embodiment, the scale factor deviation and non-calibration deviation of the accelerometer are ignored. The measurement model of the accelerometer is shown in formula (5),
其中，ak表示真实的气动加速度，bak表示加速度计偏差，ηak表示测量噪声，且whereak is the true aerodynamic acceleration,bak is the accelerometer bias,ηak is the measurement noise, and
探测器表面的压力分布可以由其携带的火星进入大气数据系统(Marsentryatmospheric data system,MEADS)实时测定，MEADS由一系列压力传感器组成，不同位置的传感器测量得到的压力与总压的关系可由公式(7)确定，The pressure distribution on the surface of the probe can be measured in real time by the Mars Entry Atmospheric Data System (MEADS) carried by the probe. MEADS is composed of a series of pressure sensors. The relationship between the pressure measured by the sensors at different positions and the total pressure can be calculated by the formula ( 7) OK,
式中，R表示静压与总压比，表示诱导角，由测压单元在探测器上位置的轴线方向与进入速度的方向确定，如图1所示，可推导where R is the ratio of static pressure to total pressure, Represents the induction angle, which is determined by the axis direction of the position of the load cell on the detector and the direction of the entry velocity, as shown in Figure 1, it can be deduced
通过牛顿流体模型，可得第i个压力传感器与总压的关系Through the Newtonian fluid model, the relationship between the ith pressure sensor and the total pressure can be obtained
其中qk,i表示真实的动压值，υq,k表示测量噪声。Where qk,i represents the real dynamic pressure value, and υq,k represents the measurement noise.
步骤二、建立进入段摄动模型。Step 2: Establish an entry-segment perturbation model.
由进入段动力学系统可知，火星大气密度及探测器气动力系数ρ与动力学系统中气动力模型紧密相关。而在组合导航方法状态估计中，不确定参数引起的摄动会随着动力学方程递推传播到下一时刻的预估状态及误差协方差中，从而导致量测更新增益出现偏差，不能得到最优状态估计。而在火星大气进入段，大气密度ρ及弹道系数B随时间存在较大的不确定性，由公式(2)、(3)知，火星大气密度及气动力系数耦合在气动力模型中，为了处理不确定参数引起的模型摄动对导航状态估计的影响，把大气密度ρ和弹道系数B对动力学的摄动归结于参数τ定义如下，From the dynamic system of the entry stage, it can be known that the density of the Martian atmosphere and the aerodynamic coefficient ρ of the probe are closely related to the aerodynamic model in the dynamic system. However, in the state estimation of the integrated navigation method, the perturbation caused by the uncertain parameters will be recursively propagated to the estimated state and error covariance at the next moment along with the dynamic equation, resulting in the deviation of the measurement update gain, which cannot be obtained. optimal state estimation. In the entry stage of the Martian atmosphere, the atmospheric density ρ and the ballistic coefficient B have greater uncertainty with time. From formulas (2) and (3), it is known that the Martian atmospheric density and aerodynamic coefficient are coupled in the aerodynamic model. In order to To deal with the influence of model perturbation caused by uncertain parameters on the estimation of navigation state, the perturbation of atmospheric density ρ and ballistic coefficient B on dynamics is attributed to the parameter τ, which is defined as follows:
大气密度ρ和弹道系统不确定性对参数τ的影响可由式(11)表示，The influence of atmospheric density ρ and ballistic system uncertainty on parameter τ can be expressed by Eq. (11),
公式(11)中，右上角带*的变量表示机载模型参数，为定值；Δ表示机载模型与真实模型的偏差，在实际飞行过程中是未知时变的，但变化范围假设为已知的。因此，大气密度ρ及弹道系数B不确定性对动力学系统的摄动在气动力模型中的体现公式(12)和公式(13)。In formula (11), the variable with * in the upper right corner represents the airborne model parameter, which is a fixed value; Δ represents the deviation between the airborne model and the real model, which is unknown and time-varying in the actual flight process, but the variation range is assumed to be Known. Therefore, the perturbation of the atmospheric density ρ and the uncertainty of the ballistic coefficient B to the dynamic system is embodied in the aerodynamic model by formula (12) and formula (13).
由公式(12)和公式(13)可知，把不确定参数大气密度ρ及弹道系数B引起的摄动转换为动力学系统中气动力的模型偏差，能够减少动力学系统中不确定项的个数，且气动力偏差的变化规律与不确定参数变化规律相同。From formula (12) and formula (13), it can be known that converting the perturbation caused by the uncertain parameters atmospheric density ρ and ballistic coefficient B into the model deviation of the aerodynamic force in the dynamic system can reduce the number of uncertain items in the dynamic system. and the change law of the aerodynamic deviation is the same as the change law of the uncertain parameters.
步骤三、利用自适应估计方法抑制不确定参数对动力学系统的扰动，提高位置和速度估计精度。Step 3: Use the adaptive estimation method to suppress the disturbance of the uncertain parameters to the dynamic system, and improve the position and velocity estimation accuracy.
根据步骤二中建立的模型摄动的有界性及其变化规律，建立针对动力学系统中摄动量Δτ的滤波模型，所述的摄动量Δτ变化范围固定且有界，每个滤波模型动力学系统中摄动量Δτ对应于有界范围内的一个偏差值，根据每个滤波模型预测测量残差，并基于测量残差信息自适应更新各模型的权值，最大权值对应的滤波模型表示该模型中摄动值最接近真实的动力学系统的摄动值。并对火星大气进入段测量信息进行无量纲化处理，无量纲化处理后的自适应估计方法各模型权值的求取方法如公式(14)所示，According to the boundedness of the model perturbation established in step 2 and its variation rule, a filter model for the perturbationΔτ in the dynamic system is established. The variation range of the perturbationΔτ is fixed and bounded. Each filter model In the dynamic system, the perturbationΔτ corresponds to a deviation value within a bounded range. The measurement residual is predicted according to each filtering model, and the weights of each model are adaptively updated based on the measurement residual information. The filter corresponding to the maximum weight The model represents the perturbation value in the model that is closest to the real dynamic system. The measurement information of the entry section of the Martian atmosphere is subjected to dimensionless processing. The method for calculating the weights of each model of the adaptive estimation method after the dimensionless processing is shown in formula (14).
权值满足公式(15)所述的条件，weight Satisfying the condition described in Equation (15),
式中，表示k时刻的无量纲化后的测量信息，Z0为已知常量，ai表示第i个滤波模型测量信息对应的输入权重。标量表示改进后第i个滤波模型与当前测量信息的匹配度。In the formula, Represents the dimensionless measurement information at time k, Z0 is a known constant, and ai represents the input weight corresponding to the measurement information of the ith filter model. scalar Indicates the matching degree of the ith filtering model after improvement with the current measurement information.
无量纲化处理后各模型测量信息权重更新方法如公式(16)所示，After dimensionless processing, the update method of the measurement information weight of each model is shown in formula (16),
式中，η表示学习律，根据实际情况由使用者自定义，hi为，In the formula, η represents the learning law, which is defined by the user according to the actual situation, andhi is,
第i个滤波模型的后验概率密度函数，且有The posterior probability density function of the ith filter model, and has
式中：where:
根据步骤一得到动力学模型和测量模型，以及步骤二中不确定参数引起的摄动模型，通过本实施例的自适应估计方法对探测器的位置r和速度v进行解算，整个方法实施过程如图2所示。本实施例采用5个滤波模型来描述不确定参数引起的摄动变化，每个滤波模型中动力学系统中气动力模型偏差各不相同，且每个滤波模型采用EKF来解算探测器的位置r和速度v。且根据火星大气进入过程的实际飞行情况，可适当增减滤波模型数量，本案例假设大气密度及弹道系数不确定性引起的摄动服从正太分布，且有Δτ～N(-0.15,0.15)，各滤波模型动力学系统对应的摄动量如表1所示，仿真中探测器初始状态及对应的偏差如表2所示，探测器构型及气动力系数如表3所示，压力传感器测量偏差及IMU中加速度噪声如表4所示，进行1000次蒙特卡洛仿真，开伞点各状态估计误差均方根如表5所示。According to the dynamic model and measurement model obtained in step 1, and the perturbation model caused by the uncertain parameters in step 2, the position r and velocity v of the detector are solved by the adaptive estimation method of this embodiment, and the whole method is implemented. as shown in picture 2. In this embodiment, five filter models are used to describe the perturbation changes caused by uncertain parameters. The deviations of the aerodynamic models in the dynamic system in each filter model are different, and each filter model uses EKF to calculate the position of the detector. r and velocity v. And according to the actual flight situation during the entry of the Martian atmosphere, the number of filter models can be appropriately increased or decreased. In this case, it is assumed that the perturbation caused by the uncertainty of the atmospheric density and ballistic coefficient obeys the normal distribution, and has Δτ ~N(-0.15,0.15) , the perturbation corresponding to each filter model dynamic system is shown in Table 1, the initial state of the detector and the corresponding deviation in the simulation are shown in Table 2, the detector configuration and aerodynamic coefficient are shown in Table 3, and the pressure sensor measurement The deviation and acceleration noise in the IMU are shown in Table 4. 1000 times of Monte Carlo simulations are performed, and the root mean square of each state estimation error at the parachute point is shown in Table 5.
表1.每个动力学模型对应的摄动偏差Table 1. Perturbation deviations corresponding to each kinetic model
表2.探测器初始状态及对应偏差Table 2. Detector initial state and corresponding deviation
表3.探测构型及气动参数Table 3. Probe configuration and aerodynamic parameters
表4.敏感器测量精度Table 4. Sensor Measurement Accuracy
表5.开伞点位置与速度估计偏差均方根误差(RMES)Table 5. Root Mean Square Error (RMES) of Open Point Position and Velocity Estimated Bias
表5给出了模型摄动下该自适应估计方法应用于火星大气进入段动压测量辅助的组合导航方法中获得的开伞点位置和速度估计误差均方差。由表5可知，该估计方法可以保证导航系统在不确定参数摄动存在的条件下三轴位置偏差在600m以内，三轴速度偏差在0.2m/s以内。从图3可以看出，本实施例所获得的状态估计偏差随进入时间逐渐收敛。Table 5 shows the mean square error of the position of the parachute point and the estimated error of the velocity obtained by the adaptive estimation method applied to the integrated navigation method assisted by the dynamic pressure measurement of the Martian atmosphere entry under the model perturbation. It can be seen from Table 5 that this estimation method can ensure that the three-axis position deviation of the navigation system is within 600m and the three-axis velocity deviation is within 0.2m/s under the condition of uncertain parameter perturbation. It can be seen from FIG. 3 that the state estimation deviation obtained in this embodiment gradually converges with the entry time.
以上所述的具体描述，对发明的目的、技术方案和有益效果进行了进一步详细说明，所应理解的是，以上所述仅为本发明的具体实施例而已，并不用于限定本发明的保护范围，凡在本发明的精神和原则之内，所做的任何修改、等同替换、改进等，均应包含在本发明的保护范围之内。The above-mentioned specific descriptions further describe the purpose, technical solutions and beneficial effects of the present invention in detail. It should be understood that the above-mentioned descriptions are only specific embodiments of the present invention, and are not intended to limit the protection of the present invention. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention shall be included within the protection scope of the present invention.
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201611242141.XACN106525055B (en) | 2016-12-29 | 2016-12-29 | An Adaptive Estimation Method for Martian Atmospheric Entry Based on Model Perturbation |
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201611242141.XACN106525055B (en) | 2016-12-29 | 2016-12-29 | An Adaptive Estimation Method for Martian Atmospheric Entry Based on Model Perturbation |
| Publication Number | Publication Date |
|---|---|
| CN106525055A CN106525055A (en) | 2017-03-22 |
| CN106525055Btrue CN106525055B (en) | 2019-04-30 |
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201611242141.XAActiveCN106525055B (en) | 2016-12-29 | 2016-12-29 | An Adaptive Estimation Method for Martian Atmospheric Entry Based on Model Perturbation |
| Country | Link |
|---|---|
| CN (1) | CN106525055B (en) |
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107480347A (en)* | 2017-07-24 | 2017-12-15 | 湖北航天技术研究院总体设计所 | A kind of chorista dispersion characteristic predicting method |
| CN109284569B (en)* | 2018-10-15 | 2023-04-07 | 天津大学 | Combined instability analysis and prediction method for periodic stator of permanent magnet motor |
| CN109612472B (en)* | 2019-01-11 | 2020-08-25 | 中国人民解放军国防科技大学 | Construction method and device for autonomous navigation system of deep space probe |
| CN110132281B (en)* | 2019-05-21 | 2023-10-20 | 哈尔滨工程大学 | Underwater high-speed target high-precision autonomous acoustic navigation method based on inquiry response mode |
| CN114019792B (en)* | 2021-10-08 | 2023-08-01 | 北京控制工程研究所 | Mars atmosphere entry process lift force guidance error analysis method and system |
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103323009A (en)* | 2013-07-10 | 2013-09-25 | 北京航空航天大学 | Non-linear three-step filtering method for Mars atmosphere entry section |
| CN103453907A (en)* | 2013-09-06 | 2013-12-18 | 北京理工大学 | Planet entering section navigation filtering method based on layered atmosphere model |
| CN104635488A (en)* | 2014-12-12 | 2015-05-20 | 北京理工大学 | Mars atmosphere entrance section robustness forecast guidance method |
| CN104794323A (en)* | 2014-12-23 | 2015-07-22 | 北京理工大学 | Martian atmosphere entrance interruption allowing estimation method based on multiple models |
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103323009A (en)* | 2013-07-10 | 2013-09-25 | 北京航空航天大学 | Non-linear three-step filtering method for Mars atmosphere entry section |
| CN103453907A (en)* | 2013-09-06 | 2013-12-18 | 北京理工大学 | Planet entering section navigation filtering method based on layered atmosphere model |
| CN104635488A (en)* | 2014-12-12 | 2015-05-20 | 北京理工大学 | Mars atmosphere entrance section robustness forecast guidance method |
| CN104794323A (en)* | 2014-12-23 | 2015-07-22 | 北京理工大学 | Martian atmosphere entrance interruption allowing estimation method based on multiple models |
| Title |
|---|
| Multimodel navigation with gating networks for Mars entry precision landing;Olivier Dubois-Matra，et al.;《AIAA Atmospheric Flight Mechanics Conference and Exhibit》;20040819;1-16 |
| 基于多模型自适应估计的火星大气层进入段导航方法;伍凯等;《中国宇航学会深空探测技术专业委员会第九届学术年会论文集》;20121231;281-288 |
| Publication number | Publication date |
|---|---|
| CN106525055A (en) | 2017-03-22 |
| Publication | Publication Date | Title |
|---|---|---|
| CN106525055B (en) | An Adaptive Estimation Method for Martian Atmospheric Entry Based on Model Perturbation | |
| CN111798491B (en) | A Maneuvering Target Tracking Method Based on Elman Neural Network | |
| US20200183339A1 (en) | Systems and Methods for Robust Learning-Based Control During Forward and Landing Flight Under Uncertain Conditions | |
| CN105973238B (en) | A kind of attitude of flight vehicle method of estimation based on norm constraint volume Kalman filtering | |
| CN108595756B (en) | Method and device for large envelope flight interference estimation | |
| CN113670314A (en) | UAV attitude estimation method based on PI adaptive two-stage Kalman filter | |
| Zamani | Deterministic attitude and pose filtering, an embedded Lie groups approach | |
| CN103218482B (en) | The method of estimation of uncertain parameter in a kind of dynamic system | |
| Lungu et al. | Application of H2/H∞ technique to aircraft landing control | |
| CN105865446A (en) | Inertia altitude channel damping Kalman filtering method based on atmosphere assistance | |
| Jin et al. | Development and validation of linear covariance analysis tool for atmospheric entry | |
| Kokunko et al. | Synthesis of a tracking system with restrictions on UAV state variables | |
| Wang et al. | Nonlinear multiple integrator and application to aircraft navigation | |
| CN105865441A (en) | Composite layered adaptive filter for multi-source disturbance system | |
| Brezoescu et al. | Wind estimation for accurate airplane path following applications | |
| CN114035536A (en) | A Robustness Evaluation Method of Flight Control System Based on Monte Carlo Method | |
| CN109857127A (en) | The method, apparatus that training neural network model and attitude of flight vehicle resolve | |
| CN113867143A (en) | Extraterrestrial celestial body safety soft landing analysis obstacle avoidance guidance method | |
| Holtsov et al. | Comparison of the UAV adaptive control with the robust control based on mu-synthesis | |
| Pourtakdoust et al. | An adaptive unscented Kalman filter for quaternion‐based orientation estimation in low‐cost AHRS | |
| Hu et al. | Hierarchic controllability analysis in high-dynamic guidance for autonomous vehicle landing | |
| Liu et al. | On terrain-aided navigation for unmanned aerial vehicle using B-spline neural network and extended Kalman filter | |
| Grillo et al. | Wind component estimation for UAS flying in turbulent air | |
| CN112947523B (en) | Angular constraint guidance method and system based on nonsingular rapid terminal sliding mode control | |
| CN108562293B (en) | A finite-time control method for planetary landing based on disturbance observer |
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | 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 |