技术领域Technical Field
本发明属于卫星定位设备信号处理技术领域，具体涉及一种用于天线内置定位终端的信号质量评估方法。The invention belongs to the technical field of satellite positioning equipment signal processing, and in particular relates to a signal quality evaluation method for a positioning terminal with a built-in antenna.
背景技术Background Art
随着全球导航卫星系统(Global Navigation Satellite System,GNSS)，包括美国的全球定位系统(Global Positioning System，GPS)、中国的北斗卫星导航系统(BeiDouNavigation Satellite System，BDS)、俄罗斯的全球卫星导航系统(Global NavigationSatellite System，GLONASS)和欧洲的伽利略卫星定位系统(Galileo SatelliteNavigation System，Galileo)等多星座GNSS的快速发展，导航定位精度显著提高。目前，随着更多智慧城市的需求和更多智能化设备的出现，GNSS成为其中不可或缺的一环，但同时也对这些定位设备的精度提出了一定的要求。为了衡量这些定位设备的性能，确定观测值的精度是非常重要的一步，否则，用户无法获得最优的定位解及其精度。With the rapid development of the Global Navigation Satellite System (GNSS), including the United States' Global Positioning System (GPS), China's BeiDou Navigation Satellite System (BDS), Russia's Global Navigation Satellite System (GLONASS) and Europe's Galileo Satellite Navigation System (Galileo), the accuracy of navigation and positioning has been significantly improved. At present, with the demand for more smart cities and the emergence of more intelligent devices, GNSS has become an indispensable part, but at the same time, it has also put forward certain requirements for the accuracy of these positioning devices. In order to measure the performance of these positioning devices, it is very important to determine the accuracy of the observation value, otherwise, users cannot obtain the optimal positioning solution and its accuracy.
近些年来，为实现准确评价定位设备的精度，使用精确的函数模型和随机模型是非常重要的。与更严谨统一的函数模型相比，随机模型的方差和协方差元素往往是未知的，因为它们与很多内部和外部影响因素有关。随机模型由方差-协方差矩阵表示，该矩阵描述了GNSS信号随机误差的期望和离散度，任何不切实际的随机模型都会对定位设备的精度造成不利影响。因此，需要通过使用精确的方法来确定方差因子，方差函数和协方差函数。通常采用基于零基线和短基线的方法来完成，即涉及到两台定位设备。在零基线法中，两台定位设备(通常型号相同)连接到同一根天线(包括前置放大器)，且基线长为零。此时，外部环境误差，包括大气延迟、多路径效应等都将被定位设备间的单差观测方程消去，且只剩下信号噪声。然而，GNSS定位设备，特别是低成本的定位设备，通常会嵌入多个天线。因此，零基线的缺点是只能评价定位设备内部信号的噪声，而无法将天线等因素考虑进去，因此得到的结果是过于乐观的。在短基线方法中，两台定位设备连接到两根天线，且天线相互之间的距离非常短，通常只有几米。因此，相同的外界环境误差被显著消除了。与零基线法不同的是，由于该方法使用了两根天线，因此得到的信号噪声来源包含了整个定位设备(包括接收机和天线等)。事实上，这也是为何零基线法计算得到的观测值精度值要小于短基线法中计算得到的观测值精度值的一个重要原因。因此，短基线法可作为零基线法的一个补充。与零基线法进行比较，短基线法存在一个明显缺陷，即观测值中会残留多路径效应等误差，且无法忽略。此外，伪距和相位观测值的随机评估方法在单频条件下效果不佳，特别是独立模式的无几何模型。并且，大多数低成本的定位设备，特别是智能手机，只能获得单频信号，而且非模型化误差的影响通常很严重。因此，为了提取纯净的观测值噪声，往往需要使用额外的数据处理方法。In recent years, it is very important to use accurate function models and random models to accurately evaluate the accuracy of positioning devices. Compared with more rigorous and unified function models, the variance and covariance elements of random models are often unknown because they are related to many internal and external influencing factors. The random model is represented by the variance-covariance matrix, which describes the expectation and dispersion of the random error of the GNSS signal. Any unrealistic random model will have an adverse effect on the accuracy of the positioning device. Therefore, it is necessary to determine the variance factor, variance function and covariance function by using accurate methods. It is usually done by a method based on zero baseline and short baseline, that is, two positioning devices are involved. In the zero baseline method, two positioning devices (usually of the same model) are connected to the same antenna (including preamplifier) and the baseline length is zero. At this time, external environmental errors, including atmospheric delay, multipath effect, etc., will be eliminated by the single difference observation equation between the positioning devices, and only signal noise will remain. However, GNSS positioning devices, especially low-cost positioning devices, usually have multiple antennas embedded. Therefore, the disadvantage of the zero baseline is that it can only evaluate the noise of the internal signal of the positioning device, but cannot take factors such as the antenna into account, so the result is too optimistic. In the short baseline method, two positioning devices are connected to two antennas, and the distance between the antennas is very short, usually only a few meters. Therefore, the same external environment error is significantly eliminated. Unlike the zero baseline method, since this method uses two antennas, the signal noise source obtained includes the entire positioning device (including receivers and antennas, etc.). In fact, this is also an important reason why the accuracy of the observation value calculated by the zero baseline method is smaller than the accuracy of the observation value calculated in the short baseline method. Therefore, the short baseline method can be used as a supplement to the zero baseline method. Compared with the zero baseline method, the short baseline method has an obvious defect, that is, errors such as multipath effects will remain in the observation value and cannot be ignored. In addition, the random evaluation method of pseudorange and phase observations does not work well under single-frequency conditions, especially the geometric model-free independent mode. Moreover, most low-cost positioning devices, especially smartphones, can only obtain single-frequency signals, and the impact of non-modeled errors is usually serious. Therefore, in order to extract pure observation noise, additional data processing methods are often required.
目前很多定位设备的天线都是内置的，无法通过相对定位模式对其进行质量评估，很有必要发展一种稳定成熟的适合评价单个定位设备且不考虑跟踪频率数量的噪声及其精度的分析方法。Currently, the antennas of many positioning devices are built-in, and their quality cannot be evaluated through relative positioning mode. It is necessary to develop a stable and mature analysis method that is suitable for evaluating the noise and accuracy of a single positioning device without considering the number of tracking frequencies.
发明内容Summary of the invention
本发明的目的在于克服现有技术中的不足之处，提供一种用于天线内置定位终端的信号质量评估方法。The purpose of the present invention is to overcome the deficiencies in the prior art and provide a signal quality evaluation method for a positioning terminal with a built-in antenna.
为了实现本发明的目的，我们将采用如下所述的技术方案加以实施。In order to achieve the purpose of the present invention, we will adopt the following technical solution to implement it.
一种用于天线内置定位终端的信号质量评估方法，所述方法包括如下所述的步骤：A signal quality assessment method for a terminal with built-in antenna positioning, the method comprising the following steps:
获取天线内置卫星定位设备的信号观测值，包括伪距和相位观测值；Obtain signal observation values of the satellite positioning device built into the antenna, including pseudorange and phase observation values;
建立顾及几何项的原始观测模型，具体步骤如下：Establish the original observation model taking into account the geometric terms. The specific steps are as follows:
天线内置卫星定位设备的伪距和相位原始观测模型如下：The pseudorange and phase original observation models of the antenna's built-in satellite positioning device are as follows:
式中，P和φ表示伪距和相位观测值；下标k和i分别表示定位设备和频率的索引；上标p表示卫星的索引；ρ表示卫地距；s表示真空中的光速；δtk和δtp分别表示定位设备和卫星的钟差；λ表示波长；N表示整周模糊度；α和β分别表示伪距和相位的硬件延迟；I和T分别表示电离层和对流层延迟；Z和z分别表示伪距和相位多路径效应；ε和∈分别表示伪距和相位的观测噪声；其他系统误差，包括相位中心的偏移和变化、相位缠绕、相对论效应、地球固体潮、海潮负荷和地球自转，都假定事先已改正；Where P and φ represent the pseudorange and phase observations; subscripts k and i represent the indexes of the positioning device and frequency, respectively; superscript p represents the index of the satellite; ρ represents the satellite-to-earth distance; s represents the speed of light in vacuum; δtk and δtp represent the clock errors of the positioning device and satellite, respectively; λ represents the wavelength; N represents the integer ambiguity; α and β represent the hardware delays of the pseudorange and phase, respectively; I and T represent the ionospheric and tropospheric delays, respectively; Z and z represent the multipath effects of the pseudorange and phase, respectively; ε and ∈ represent the observation noise of the pseudorange and phase, respectively; other systematic errors, including the shift and change of the phase center, phase entanglement, relativistic effects, the earth's solid tide, ocean tide loads, and the earth's rotation, are assumed to have been corrected in advance;
因顾及几何项，所以可以采用基于几何固定、基于几何或基于无几何三种方式去建立原始观测模型。这里以基于无几何为例，对于上述原始观测模型，可以通过参数重组进行变换，具体为：Taking the geometric terms into consideration, the original observation model can be established in three ways: based on fixed geometry, based on geometry, or based on no geometry. Here, taking the no geometry as an example, the original observation model can be transformed by parameter reorganization, specifically:
式中，等效卫地距表示非分散项，等效伪距多路径包含伪距多路径本身，以及定位设备端和卫星端的硬件延迟，等效多路径包含相位多路径本身、整周模糊度以及定位设备端和卫星端的硬件延迟；对应的紧凑形式如下：In the formula, the equivalent satellite-to-ground distance is represents the non-dispersive term, equivalent pseudorange multipath Includes pseudorange multipath itself, as well as hardware delays on the positioning device and satellite, equivalent multipath It includes the phase multipath itself, integer ambiguity, and hardware delays of the positioning device and the satellite. The corresponding compact form is as follows:
式中，cf＝[c1,…,ci]T；其中电离层延迟系数cj等于In the formula, cf =[c1 ,…,ci ]T ; The ionospheric delay coefficientcj is equal to
因此时上述观测模型存在秩亏问题，无法求解观测模型，且在多历元模式下也存在问题，所以为了避免秩亏，先对电离层延迟进行相应的改正。电离层延迟需要提前进行更好的校正，可以使用一些外部电离层经验模型或电离层产品进行改正，例如Klobuchar模型等外部电离层产品。无偏模型，如电离层固定、电离层浮点或电离层加权模型也可以应用。Since the above observation model has a rank deficiency problem, the observation model cannot be solved, and there are also problems in the multi-epoch mode, so in order to avoid rank deficiency, the ionospheric delay is corrected accordingly. The ionospheric delay needs to be better corrected in advance, and some external ionospheric empirical models or ionospheric products can be used for correction, such as external ionospheric products such as the Klobuchar model. Unbiased models such as ionospheric fixed, ionospheric floating point or ionospheric weighted models can also be applied.
基于上述基于无几何函数模型的观测模型，建立考虑非模型化误差的观测模型，具体步骤如下：Based on the above observation model based on the geometry-free model, an observation model considering non-modeled errors is established. The specific steps are as follows:
假设电离层误差在上一步得到改正，将残余电离层延迟和等效多路径效应重新组合成一个新的参数，称为等效非模型化误差，因此，观测模型可以更进一步改进为：Assuming that the ionospheric error is corrected in the previous step, the residual ionospheric delay and the equivalent multipath effect are recombined into a new parameter called the equivalent unmodeled error. Therefore, the observation model can be further improved as follows:
式中，伪距非模型化误差和相位非模型化误差其中，dI表示可能存在的残余电离层延迟；对应的紧凑形式如下：Where, the pseudorange non-modeled error is and phase unmodeled error where dI represents the possible residual ionospheric delay; the corresponding compact form is as follows:
式中，Ψ＝[WT,wT]T,由于上述紧凑观测模型秩亏，且要使得参数可估计，因此需要选择为消除秩亏而作为基准的参数，对观测模型进行重新参数化。主要步骤如下：Where, Ψ＝[WT ,WT ]T , Since the above compact observation model is rank-deficient and the parameters need to be estimable, it is necessary to select parameters that are used as a benchmark to eliminate the rank deficiency and reparameterize the observation model. The main steps are as follows:
频率m上的伪距非模型化误差Wm参数被选为基准，具体来说，将设计矩阵Wm改为0，从而Wm使被其他伪距和相位非模型化误差吸收。可选取第一频率上的伪距非模型化误差作为消除秩亏的基准。重新参数化的观测模型具体为：The pseudorange non-modeled errorWm parameter at frequency m is selected as the benchmark. Specifically, the design matrixWm is changed to 0 so thatWm is absorbed by other pseudorange and phase non-modeled errors. The pseudorange non-modeled error at the first frequency can be selected as the benchmark for eliminating rank deficiency. The re-parameterized observation model is specifically:
式中，Γn,2fn＝blkdiag(0n,I(2f-1)n)，运算符‘blkdiag’表示矩阵对角拼接计算；没有其中等效伪距和相位非模型化误差满足和此时，设计矩阵满足列满秩，可以进行参数估计。Wherein, Γn,2fn = blkdiag(0n ,I(2f-1)n ), the operator 'blkdiag' represents the matrix diagonal concatenation calculation; No The equivalent pseudorange and phase non-modeled errors satisfy and At this point, the design matrix satisfies the full column rank and parameter estimation can be performed.
因没有冗余测量，所以需要合适的处理等效非模型化误差，因此联立多历元观测值求解等效非模型化误差。具体步骤如下：Since there is no redundant measurement, it is necessary to properly handle the equivalent non-modeled error, so the equivalent non-modeled error is solved by combining multiple epoch observations. The specific steps are as follows:
因单历元条件下缺少多余观测值，所以需要合适的处理等效非模型化误差。在等效伪距和相位非模型化误差中，如果没有发生周跳，接收机端和卫星端的硬件延迟和整周模糊度(仅相位有整周模糊度)可视作常数。因多路径和残余大气延迟在短时间内具有较高时间相关性，所以需要合适地处理多路径和残余大气延迟。对于伪距等效非模型化误差，第一频率上仍然存在伪距非模型化误差、剩余电离层延迟、伪距多路径效应和伪距硬件延迟。相位等效非模型化误差中包含了第一频率上的伪距非模型化误差、剩余电离层延迟、相位多路径效应、相位硬件延迟和整周模糊度。Due to the lack of redundant observations under single epoch conditions, it is necessary to properly handle the equivalent non-modeled errors. In the equivalent pseudorange and phase non-modeled errors, if there is no cycle slip, the hardware delay and integer ambiguity (only the phase has integer ambiguity) at the receiver and satellite ends can be regarded as constants. Because multipath and residual atmospheric delay have high time correlation in a short period of time, it is necessary to properly handle multipath and residual atmospheric delay. For the pseudorange equivalent non-modeled error, there are still pseudorange non-modeled errors, residual ionospheric delay, pseudorange multipath effect and pseudorange hardware delay on the first frequency. The phase equivalent non-modeled error includes the pseudorange non-modeled error, residual ionospheric delay, phase multipath effect, phase hardware delay and integer ambiguity on the first frequency.
因等效非模型化误差在短时间内可视为时不变参数，所以可以采用多历元参数化来求解等效多路径效应，具体步骤如下：Since the equivalent unmodeled error can be regarded as a time-invariant parameter in a short period of time, multi-epoch parameterization can be used to solve the equivalent multipath effect. The specific steps are as follows:
首先设置一个移动窗口，考虑到非模型误差间的物理相关性，且需准确估计非模型化误差，所以移动窗口的大小非常重要。对于1HZ的观测数据，由于两个连续历元间的非模型化误差引起的时间相关性通常高于0.95甚至0.99，且第一次和第五次测量之间的时间相关性仍然至少高于0.80。另一方面，由于冗余测量的数量有限，如果移动窗口太小则无法准确的估计非模型误差。因此移动窗口的大小最好设置在3到5个历元之间。值得注意的是，如果采样率不是1HZ，则最优的移动窗口大小可能会改变。First, a moving window is set. Considering the physical correlation between non-model errors and the need to accurately estimate non-model errors, the size of the moving window is very important. For 1HZ observation data, the time correlation caused by non-model errors between two consecutive epochs is usually higher than 0.95 or even 0.99, and the time correlation between the first and fifth measurements is still at least higher than 0.80. On the other hand, due to the limited number of redundant measurements, if the moving window is too small, the non-model errors cannot be accurately estimated. Therefore, the size of the moving window is preferably set between 3 and 5 epochs. It is worth noting that if the sampling rate is not 1HZ, the optimal moving window size may change.
此时，可估可解的观测模型如下：At this point, the estimable and solvable observation model is as follows:
式中，Pξ＝[PT(t-ξ+1),…,PT(t)]T，其中P(t)表示在历元t处的伪距观测值；φξ＝[φT(t-ξ+1),…,φT(t)]T，其中φ(t)表示在历元t处的相位观测值；其中表示在历元t处的等效卫地距。Wherein, Pξ =[PT (t-ξ+1),…,PT (t)]T , where P(t) represents the pseudorange observation value at epoch t; φξ =[φT (t-ξ+1),…,φT (t)]T , where φ(t) represents the phase observation value at epoch t; in represents the equivalent satellite-earth distance at epoch t.
获取天线内置定位设备的信号噪声的具体步骤如下：因其他系统误差、非模型化误差都得到了很好的改正或估计，因此此时可以获取天线内置定位设备的信号噪声，即求解天线内置定位设备的伪距和相位的观测值残差。The specific steps for obtaining the signal noise of the antenna's built-in positioning device are as follows: Since other system errors and non-modeled errors have been well corrected or estimated, the signal noise of the antenna's built-in positioning device can be obtained at this time, that is, the observation value residuals of the pseudorange and phase of the antenna's built-in positioning device are solved.
利用自回归滑动平均模型求解方差因子的具体步骤如下：The specific steps of solving the variance factor using the autoregressive moving average model are as follows:
因上述伪距和相位观测值残差，即信号噪声附着具有随高度角而变化的趋势项，因此需要对信号噪声进行提纯，来获取信号的白噪声。对上述伪距或相位观测值残差序列简称为ΔG，式中，Δ代表时间差分算子，满足Δ[.](t)＝[.](t)-[.](t-1)；t代表观测历元。Because the above pseudorange and phase observation residuals, that is, the signal noise attachment, have a trend term that changes with the altitude angle, it is necessary to purify the signal noise to obtain the white noise of the signal. The above pseudorange or phase observation residual sequence is referred to as ΔG, where Δ represents the time difference operator, satisfying Δ[.](t)=[.](t)-[.](t-1); t represents the observation epoch.
ΔG包含两种多项式的平稳过程，其中ΔG(t-1),ΔG(t-2),…,ΔG(t-e)为自回归过程，而v(t-1),v(t-2),…,v(t-g)为滑动平均过程，其中v(t-1),v(t-2),…,v(t-g)均为时间差分的信号的噪声，e和g为自回归滑动平均模型的阶数。因此，ΔG满足以下条件：ΔG contains two polynomial stationary processes, where ΔG(t-1), ΔG(t-2), …, ΔG(t-e) are autoregressive processes, and v(t-1), v(t-2), …, v(t-g) are moving average processes, where v(t-1), v(t-2), …, v(t-g) are all noises of time-differenced signals, and e and g are the orders of the autoregressive moving average model. Therefore, ΔG satisfies the following conditions:
ΔG(t)＝f(t)δ+v(t)ΔG(t)＝f(t)δ+v(t)
式中，f(x)＝[ΔG(t-1),ΔG(t-2),…,ΔG(t-e),v(t-1),v(t-2),…,v(t-g)]；是自回归模型参数，是滑动平均参数；如果存在U个历元，则多历元模型具体为：In the formula, f(x)=[ΔG(t-1),ΔG(t-2),…,ΔG(te),v(t-1),v(t-2),…,v(tg) ]; are the autoregressive model parameters, is the sliding average parameter; if there are U epochs, the multi-epoch model is specifically:
Y(U)＝F(U)β+V(U)Y(U)＝F(U)β+V(U)
式中，Y(U)＝[ΔG(t),ΔG(t+1),…,ΔG(t+U-1)]T；F(U)＝[f(t),f(t+1),…,f(t+U-1)]T；V(L)＝[v(t),v(t+1),…,v(t+U-1)]T。相应的U个历元的最小二乘解为：Where, Y(U) = [ΔG(t), ΔG(t+1), …, ΔG(t+U-1)]T ; F(U) = [f(t), f(t+1), …, f(t+U-1)]T ; V(L) = [v(t), v(t+1), …, v(t+U-1)]T . The corresponding least squares solution for U epochs is:
由于f(t)依赖于之前的历元，因此该过程的估计是一个迭代的过程，其中，自回归滑动平均模型阶数e和g满足获得的BIC值最小：Since f(t) depends on the previous epoch, the estimation process is an iterative process, in which the autoregressive moving average model orders e and g satisfy The minimum BIC value is obtained:
式中，u是与参数估值对应的优化对数似然函数值；最后，得到信号的白噪声，即方差因子；Where u is the optimized log-likelihood function value corresponding to the parameter estimation; finally, the white noise of the signal is obtained, that is, the variance factor;
利用高度角模型或载噪比模型求解方差函数的具体步骤如下：The specific steps for solving the variance function using the altitude angle model or the carrier-to-noise ratio model are as follows:
方差函数可以由：The variance function can be given by:
(1)、高度角模型(1) Altitude angle model
来确定的，式中，σ表示由高度角θ估计的方差分量；a和b为最小二乘准则确定的系数。is determined by, where σ represents the variance component estimated by the altitude angle θ; a and b are coefficients determined by the least squares criterion.
(2)、载噪比模型由(2) The carrier-to-noise ratio model is given by
来确定的；式中，Bi为相位跟踪环带宽(Hz)，λ为载波相位波长(m)。Ci的取至与载波信号波长以及接收机内部的跟踪硬件有关，例如，对于GPS的L1频点以及L2频点分别取值为C1＝0.00224m2HZ，C2＝0.00077m2HZ。whereBi is the phase tracking loop bandwidth (Hz) and λ is the carrier phase wavelength (m). The value ofCi is related to the carrier signal wavelength and the tracking hardware inside the receiver. For example, the values of L1 and L2 frequencies of GPS areC1 =0.00224m2HZ andC2 =0.00077m2HZ respectively.
利用自相关或互相关技术求解协方差函数的具体步骤为：The specific steps to solve the covariance function using autocorrelation or cross-correlation techniques are:
所述的协方差函数展现了观测值之间的数学相关性和物理相关性，在物理相关性中，包括了空间相关性、时间相关性和交叉相关性。The covariance function shows the mathematical correlation and physical correlation between the observations, and the physical correlation includes spatial correlation, temporal correlation and cross-correlation.
利用自相关技术对观测值之间的时间相关性进行估计，由：The temporal correlation between observations is estimated using the autocorrelation technique, given by:
来确定的；式中，ι为时间间隔，并满足n为观测值残差个数，ζ(j)和ζ(j+ι)为第j和j+ι个历元的观测值残差，为n个观测值残差的平均值。to determine; where ι is the time interval and satisfies n is the number of observation residuals, ζ(j) and ζ(j+ι) are the observation residuals of the jth and j+ιth epochs, is the mean of the residuals of the n observations.
利用互相关技术对观测值之间的空间相关性和交叉相关性进行估计，由：The spatial correlation and cross-correlation between observations are estimated using cross-correlation techniques, given by:
来确定的；式中，γpq是观测值Op和Oq的互相关系数，γqp是观测值Oq和Op的互相关系数，θp和θq分别是观测值Op和Oq对应的残差ζp和ζq的标准差，Cov为协方差算子。is determined by; where γpq is the mutual correlation coefficient of the observations Op and Oq , γqp is the mutual correlation coefficient of the observations Oq and Op , θp and θq are the standard deviations of the residuals ζp and ζq corresponding to the observations Op and Oq, respectively, and Cov is the covariance operator.
有益效果Beneficial Effects
本发明提供的一种用于天线内置定位终端的信号质量评估方法，只需一台卫星定位设备，且无须考虑设备跟踪频率的数量即可准确可靠的确定其信号质量；The present invention provides a signal quality evaluation method for a positioning terminal with an antenna built in, which only requires one satellite positioning device and can accurately and reliably determine its signal quality without considering the number of device tracking frequencies;
本发明简化了定位设备的信号质量评估流程，本发明不需要进行零基线和短基线的测试来进行评估定位设备的信号质量，大大简化了信号质量评估的流程，降低了评估定位设备所需的成本以及对设备的要求；The present invention simplifies the signal quality evaluation process of the positioning device. The present invention does not need to perform zero baseline and short baseline tests to evaluate the signal quality of the positioning device, which greatly simplifies the signal quality evaluation process and reduces the cost and equipment requirements required for evaluating the positioning device.
本发明提高了信号质量评估的适用性和全面性，在面对一些复杂场景，可能无法进行零基线和短基线进行评估定位设备的信号质量，而本发明可以即时的在这些复杂场景进行准确的评估信号质量，可以迅速的适应不同的信号环境，提高了设备在动态环境中的性能；The present invention improves the applicability and comprehensiveness of signal quality assessment. In some complex scenarios, it may not be possible to use zero baseline and short baseline to assess the signal quality of the positioning device. However, the present invention can accurately assess the signal quality in these complex scenarios in real time, can quickly adapt to different signal environments, and improve the performance of the device in a dynamic environment.
通过在内置天线的定位设备中实施本发明的技术方案，可以更准确的评估信号质量，从而提高定位的精度和稳定性。这对于需要高精度定位的设备、应用等都具有显著的优势。By implementing the technical solution of the present invention in a positioning device with a built-in antenna, the signal quality can be evaluated more accurately, thereby improving the accuracy and stability of positioning, which has significant advantages for devices and applications that require high-precision positioning.
附图说明BRIEF DESCRIPTION OF THE DRAWINGS
图1为一种用于天线内置定位终端的信号质量评估方法的流程图。FIG1 is a flow chart of a signal quality assessment method for a positioning terminal with a built-in antenna.
具体实施方式DETAILED DESCRIPTION
为了对本发明的技术特征、目的和效果有更加清楚的理解，现依据附图和实施例对本发明做进一步详细说明。In order to have a clearer understanding of the technical features, purposes and effects of the present invention, the present invention is now further described in detail based on the accompanying drawings and embodiments.
作为本发明的实施例，如图1所示，一种用于天线内置定位终端的信号质量评估方法，包括如下步骤：As an embodiment of the present invention, as shown in FIG1 , a signal quality assessment method for a positioning terminal with a built-in antenna includes the following steps:
获取该设备的单频观测值，包括伪距和相位观测值。Get the single frequency observation values of the device, including pseudorange and phase observation values.
建立顾及几何项的原始观测模型，具体步骤如下：Establish the original observation model taking into account the geometric terms. The specific steps are as follows:
对于该天线内置的卫星定位设备，伪距和相位的原始观测模型如下：For the satellite positioning device built into the antenna, the original observation model of pseudorange and phase is as follows:
式中，P和φ表示伪距和相位观测值；下标k和i分别表示定位设备和频率的索引；上标p表示卫星的索引；ρ表示卫地距；s表示真空中的光速；δtk和δtp分别表示定位设备和卫星的钟差；λ表示波长；N表示整周模糊度；α和β分别表示伪距和相位的硬件延迟；I和T分别表示电离层和对流层延迟；Z和z分别表示伪距和相位多路径效应；ε和∈分别表示伪距和相位的观测噪声；其他系统误差，包括相位中心的偏移和变化、相位缠绕、相对论效应、地球固体潮、海潮负荷和地球自转，都假定是得到预先改正的；Where P and φ represent the pseudorange and phase observations; subscripts k and i represent the indexes of the positioning device and frequency, respectively; superscript p represents the index of the satellite; ρ represents the satellite-to-earth distance; s represents the speed of light in vacuum; δtk and δtp represent the clock errors of the positioning device and satellite, respectively; λ represents the wavelength; N represents the integer ambiguity; α and β represent the hardware delays of the pseudorange and phase, respectively; I and T represent the ionospheric and tropospheric delays, respectively; Z and z represent the pseudorange and phase multipath effects, respectively; ε and ∈ represent the observation noise of the pseudorange and phase, respectively; other systematic errors, including the shift and change of the phase center, phase entanglement, relativistic effects, earth solid tides, ocean tide loads, and earth rotation, are assumed to be corrected in advance;
基于无几何，上述原始观测模型可以通过参数重组进行变换，具体为：Based on the geometry-free model, the original observation model can be transformed by parameter reorganization, specifically:
式中，等效卫地距表示非分散项，等效伪距多路径包含伪距多路径本身，以及定位设备端和卫星端的硬件延迟，等效多路径包含相位多路径本身、整周模糊度以及定位设备端和卫星端的硬件延迟；对应的紧凑形式如下：In the formula, the equivalent satellite-to-ground distance is represents the non-dispersive term, equivalent pseudorange multipath Includes pseudorange multipath itself, as well as hardware delays on the positioning device and satellite, equivalent multipath It includes the phase multipath itself, integer ambiguity, and hardware delays of the positioning device and the satellite. The corresponding compact form is as follows:
式中，cf＝[c1,…,ci]T；其中电离层延迟系数cj等于In the formula, cf =[c1 ,…,ci ]T ; The ionospheric delay coefficientcj is equal to
因此时上述观测模型存在秩亏问题，无法求解观测模型，且在多历元模式下也存在问题，所以为了避免秩亏，先使用Klobuchar模型等外部电离层产品对电离层延迟进行相应的改正。At this time, the above observation model has a rank deficiency problem and cannot be solved. There are also problems in the multi-epoch mode. Therefore, in order to avoid rank deficiency, the external ionospheric products such as the Klobuchar model are first used to make corresponding corrections to the ionospheric delay.
基于上述基于无几何函数模型的观测模型，建立考虑非模型化误差的观测模型，具体步骤如下：Based on the above observation model based on the geometry-free model, an observation model considering non-modeled errors is established. The specific steps are as follows:
假设电离层误差得到改正，将残余电离层延迟和等效多路径效应重新组合成一个新的参数，称为等效非模型化误差，因此，观测模型可以更进一步改进为：Assuming that the ionospheric error is corrected, the residual ionospheric delay and the equivalent multipath effect are recombined into a new parameter called the equivalent unmodeled error. Therefore, the observation model can be further improved as follows:
式中，伪距非模型化误差和相位非模型化误差其中，dI表示可能存在的残余电离层延迟；相应的紧凑形式如下：Where, the pseudorange non-modeled error is and phase unmodeled error where dI is the possible residual ionospheric delay; the corresponding compact form is as follows:
式中，Ψ＝[WT,wT]T,由于上述紧凑观测模型秩亏，因此需要选择为消除秩亏而作为基准的参数，对观测模型进行重新参数化。主要步骤如下：Where, Ψ＝[WT ,WT ]T , Since the above compact observation model is rank-deficient, it is necessary to select parameters as a benchmark to eliminate the rank deficiency and reparameterize the observation model. The main steps are as follows:
频率m上的伪距非模型化误差Wm参数被选为基准，具体来说，将Wm的设计矩阵改为0，从而Wm使被其他伪距和相位非模型化误差吸收。选取GPS的C1C和BDS的C2I上的伪距非模型化误差作为消除秩差的基准。重新参数化的观测模型具体为：The pseudorange unmodeled errorWm parameter on frequency m is selected as the benchmark. Specifically, the design matrix ofWm is changed to 0, so thatWm is absorbed by other pseudorange and phase unmodeled errors. The pseudorange unmodeled errors on C1C of GPS and C2I of BDS are selected as the benchmark for eliminating rank differences. The reparameterized observation model is specifically:
式中，Γn,2fn＝blkdiag(0n,I(2f-1)n)，运算符‘blkdiag’表示矩阵对角拼接计算；没有其中等效伪距和相位非模型化误差满足和此时，设计矩阵满足列满秩，可以进行参数估计。Wherein, Γn,2fn = blkdiag(0n ,I(2f-1)n ), the operator 'blkdiag' represents the matrix diagonal concatenation calculation; No The equivalent pseudorange and phase unmodeled errors satisfy and At this point, the design matrix satisfies the full column rank and parameter estimation can be performed.
因没有冗余测量，所以需要合适的处理等效非模型化误差，在此联立多历元观测值求解等效非模型化误差。主要步骤如下：Since there is no redundant measurement, it is necessary to properly process the equivalent non-modeled error. Here, the equivalent non-modeled error is solved by combining multi-epoch observations. The main steps are as follows:
因单历元条件下缺少多余观测值，所以需要合适的处理等效非模型化误差。在等效伪距和相位非模型化误差中，如果没有发生周跳，接收机端和卫星端的硬件延迟和整周模糊度(仅相位有整周模糊度)可视作常数。因多路径和残余大气延迟在短时间内具有较高时间相关性，所以需要合适地处理多路径和残余大气延迟。且对于伪距等效非模型化误差，第一频率上仍然存在伪距非模型化误差、剩余电离层延迟、伪距多路径效应和伪距硬件延迟。相位等效非模型化误差中包含了第一频率上的伪距非模型化误差、剩余电离层延迟、相位多路径效应、相位硬件延迟和整周模糊度。Due to the lack of redundant observations under single epoch conditions, it is necessary to properly handle the equivalent non-modeled error. In the equivalent pseudorange and phase non-modeled errors, if there is no cycle slip, the hardware delay and integer ambiguity (only the phase has integer ambiguity) at the receiver and satellite ends can be regarded as constants. Because multipath and residual atmospheric delay have high time correlation in a short period of time, it is necessary to properly handle multipath and residual atmospheric delay. And for the pseudorange equivalent non-modeled error, there are still pseudorange non-modeled errors, residual ionospheric delay, pseudorange multipath effect and pseudorange hardware delay on the first frequency. The phase equivalent non-modeled error includes the pseudorange non-modeled error, residual ionospheric delay, phase multipath effect, phase hardware delay and integer ambiguity on the first frequency.
因等效非模型化误差在短时间内可视为时不变参数，所以可以使用多历元参数化来求解等效多路径效应。具体而言，设置一个移动窗口，考虑到非模型误差之间的物理相关性，且需考虑准确的估计非模型化误差，故对于该设备，选取3个历元作为移动窗口的大小最为合适。Since the equivalent non-modeled error can be regarded as a time-invariant parameter in a short time, multi-epoch parameterization can be used to solve the equivalent multipath effect. Specifically, a moving window is set, taking into account the physical correlation between the non-modeled errors and the need to accurately estimate the non-modeled errors. Therefore, for this device, 3 epochs are selected as the most appropriate size of the moving window.
此时，可估可解的观测模型如下：At this point, the estimable and solvable observation model is as follows:
式中，Pξ＝[PT(t-ξ+1),…,PT(t)]T，其中P(t)表示在历元t处的伪距观测值；φξ＝[φT(t-ξ+1),…,φT(t)]T，其中φ(t)表示在历元t处的相位观测值；其中表示在历元t处的等效卫地距。Wherein, Pξ =[PT (t-ξ+1),…,PT (t)]T , where P(t) represents the pseudorange observation value at epoch t; φξ =[φT (t-ξ+1),…,φT (t)]T , where φ(t) represents the phase observation value at epoch t; in represents the equivalent satellite-earth distance at epoch t.
获取天线内置定位设备的信号噪声的具体步骤如下：The specific steps to obtain the signal noise of the antenna built-in positioning device are as follows:
因其他系统误差、非模型化误差都得到了很好的改正或估计，因此此时可以获取天线内置定位设备的信号噪声，即求解天线内置定位设备的伪距和相位的观测值残差。所述的求解天线内置定位设备的伪距和相位的残差均为现有技术，因此本发明中不做具体赘述；具体实践中，可以采用残差＝观测值-卫地距-改正项。所述的利用自回归滑动平均模型求解方差因子的具体步骤如下：Because other system errors and non-modeled errors have been well corrected or estimated, the signal noise of the antenna built-in positioning device can be obtained at this time, that is, the residual of the observed value of the pseudorange and phase of the antenna built-in positioning device can be solved. The residuals of the pseudorange and phase of the antenna built-in positioning device are all prior art, so they are not described in detail in the present invention; in specific practice, residual = observation value - satellite-to-earth distance - correction term can be used. The specific steps of solving the variance factor using the autoregressive moving average model are as follows:
因上述伪距和相位观测值残差，即信号噪声附着具有随高度角而变化的趋势项，因此需要对信号噪声进行提纯，来获取信号的白噪声。对上述伪距或相位观测值残差序列简称为ΔG，式中，Δ代表时间差分算子，满足Δ[.](t)＝[.](t)-[.](t-1)；t代表观测历元。Because the above pseudorange and phase observation residuals, that is, the signal noise attachment, have a trend term that changes with the altitude angle, it is necessary to purify the signal noise to obtain the white noise of the signal. The above pseudorange or phase observation residual sequence is referred to as ΔG, where Δ represents the time difference operator, satisfying Δ[.](t)=[.](t)-[.](t-1); t represents the observation epoch.
ΔG包含两种多项式的平稳过程，其中ΔG(t-1),ΔG(t-2),…,ΔG(t-e)为自回归过程，而v(t-1),v(t-2),…,v(t-g)为滑动平均过程，其中v(t-1),v(t-2),…,v(t-g)均为时间差分的信号的噪声，e和g为自回归滑动平均模型的阶数。因此，ΔG满足以下条件：ΔG contains two polynomial stationary processes, where ΔG(t-1), ΔG(t-2), …, ΔG(t-e) are autoregressive processes, and v(t-1), v(t-2), …, v(t-g) are moving average processes, where v(t-1), v(t-2), …, v(t-g) are all noises of time-differenced signals, and e and g are the orders of the autoregressive moving average model. Therefore, ΔG satisfies the following conditions:
ΔG(t)＝f(t)δ+v(t)ΔG(t)＝f(t)δ+v(t)
式中，f(t)＝[ΔG(t-1),ΔG(t-2),…,ΔG(t-e),v(t-1),v(t-2),…,v(t-g)]；是自回归模型参数，是滑动平均参数；如果存在U个历元，则多历元模型具体为：In the formula, f(t)=[ΔG(t-1),ΔG(t-2),…,ΔG(te),v(t-1),v(t-2),…,v(tg) ]; are the autoregressive model parameters, is the sliding average parameter; if there are U epochs, the multi-epoch model is specifically:
Y(U)＝F(U)β+V(U)Y(U)＝F(U)β+V(U)
式中，Y(U)＝[ΔG(t),ΔG(t+1),…,ΔG(t+U-1)]T；F(U)＝[f(t),f(t+1),…,f(t+U-1)]T；V(L)＝[v(t),v(t+1),…,v(t+U-1)]T。相应的U个历元的最小二乘解为：Where, Y(U) = [ΔG(t), ΔG(t+1), …, ΔG(t+U-1)]T ; F(U) = [f(t), f(t+1), …, f(t+U-1)]T ; V(L) = [v(t), v(t+1), …, v(t+U-1)]T . The corresponding least squares solution for U epochs is:
由于f(t)依赖于之前的历元，因此该过程的估计是一个迭代的过程，其中，自回归滑动平均模型阶数e和g满足获得的BIC值最小：Since f(t) depends on the previous epoch, the estimation process is an iterative process, in which the autoregressive moving average model orders e and g satisfy The minimum BIC value is obtained:
式中，u是与参数估值对应的优化对数似然函数值；最后，得到信号的白噪声，即方差因子；Where u is the optimized log-likelihood function value corresponding to the parameter estimation; finally, the white noise of the signal is obtained, that is, the variance factor;
利用高度角模型求解方差函数，具体而言，设置一定的高度角区间，计算每个区间内观测值残差的标准差。具体由：The variance function is solved using the altitude angle model. Specifically, a certain altitude angle interval is set and the standard deviation of the residual of the observation value in each interval is calculated. Specifically:
来确定的；式中，σ表示由高度角θ估计的方差分量；a和b为最小二乘准则确定的系数。is determined by; where σ represents the variance component estimated by the altitude angle θ; a and b are coefficients determined by the least squares criterion.
利用自相关或互相关技术求解协方差函数的具体步骤为：The specific steps to solve the covariance function using autocorrelation or cross-correlation techniques are:
所述的协方差函数展现了观测值之间的数学相关性和物理相关性，在物理相关性中，包括了空间相关性、时间相关性和交叉相关性。The covariance function shows the mathematical correlation and physical correlation between the observations, and the physical correlation includes spatial correlation, temporal correlation and cross-correlation.
利用自相关技术对观测值之间的时间相关性进行估计，由：The temporal correlation between observations is estimated using the autocorrelation technique, given by:
来确定的；式中，ι为时间间隔，并满足n为观测值残差个数，ζ(j)和ζ(j+ι)为第j和j+ι个历元的观测值残差，为n个观测值残差的平均值。to determine; where ι is the time interval and satisfies n is the number of observation residuals, ζ(j) and ζ(j+ι) are the observation residuals of the jth and j+ιth epochs, is the mean of the residuals of the n observations.
利用互相关技术对观测值之间的空间相关性和交叉相关性进行估计，由：The spatial correlation and cross-correlation between observations are estimated using cross-correlation techniques, given by:
来确定的；式中，γpq是观测值Op和Oq的互相关系数，γqp是观测值Oq和Op的互相关系数，θp和θq分别是观测值Op和Oq对应的残差ζp和ζq的标准差，Cov为协方差算子。is determined by; where γpq is the mutual correlation coefficient of the observations Op and Oq , γqp is the mutual correlation coefficient of the observations Oq and Op , θp and θq are the standard deviations of the residuals ζp and ζq corresponding to the observations Op and Oq, respectively, and Cov is the covariance operator.
尽管已经示出和描述了本发明的实施例，对于本领域的普通技术人员而言，可以理解在不脱离本发明的原理和精神的情况下可以对这些实施例进行多种变化、修改、替换和变型，本发明的范围由所附权利要求及其等同物限定。Although embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions and variations may be made to the embodiments without departing from the principles and spirit of the present invention, and that the scope of the present invention is defined by the appended claims and their equivalents.
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202410146484.4ACN118011434B (en) | 2024-02-02 | 2024-02-02 | A signal quality evaluation method for a positioning terminal with built-in antenna |
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202410146484.4ACN118011434B (en) | 2024-02-02 | 2024-02-02 | A signal quality evaluation method for a positioning terminal with built-in antenna |
| Publication Number | Publication Date |
|---|---|
| CN118011434A CN118011434A (en) | 2024-05-10 |
| CN118011434Btrue CN118011434B (en) | 2024-10-01 |
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202410146484.4AActiveCN118011434B (en) | 2024-02-02 | 2024-02-02 | A signal quality evaluation method for a positioning terminal with built-in antenna |
| Country | Link |
|---|---|
| CN (1) | CN118011434B (en) |
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110703288A (en)* | 2019-10-15 | 2020-01-17 | 河海大学 | Signal quality evaluation method suitable for satellite positioning equipment with built-in antenna |
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108061911A (en)* | 2018-01-16 | 2018-05-22 | 东南大学 | A kind of GLONASS carrier waves list difference residual error method of estimation |
| CN111596322B (en)* | 2020-05-20 | 2022-12-23 | 中国电子科技集团公司第二十研究所 | A Consistency Evaluation Method for Receiver Pseudorange Measurements under Short Baseline Conditions |
| CN114488228B (en)* | 2022-04-11 | 2022-07-01 | 南京北斗创新应用科技研究院有限公司 | GNSS multi-path error weakening method suitable for dynamic carrier platform |
| CN115267843B (en)* | 2022-06-14 | 2023-05-30 | 中国科学院精密测量科学与技术创新研究院 | Real-time non-difference estimation method for clock difference of multi-frequency multi-mode GNSS high-frequency precise satellite |
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110703288A (en)* | 2019-10-15 | 2020-01-17 | 河海大学 | Signal quality evaluation method suitable for satellite positioning equipment with built-in antenna |
| Title |
|---|
| "Unmodeled-error-corrected stochastic assessment for a standalone GNSS receiver regardless of the number of tracked frequencies";Zhetao Zhang等;《Measurement》;20221128;第1-11页* |
| Publication number | Publication date |
|---|---|
| CN118011434A (en) | 2024-05-10 |
| Publication | Publication Date | Title |
|---|---|---|
| CN109738917B (en) | Multipath error weakening method and device in Beidou deformation monitoring | |
| RU2354991C2 (en) | Method of using three gps frequencies for resolving integral uncertainties of carrier phases | |
| US10330795B2 (en) | GNSS positioning system and method using multiple processing threads | |
| Su et al. | Triple-frequency carrier phase precise time and frequency transfer models for BDS-3 | |
| CN106569242B (en) | The mono- poor processing method of the GNSS of fixed reference star | |
| CN108873029B (en) | Method for realizing clock error modeling of navigation receiver | |
| CN107728171B (en) | Particle filter based real-time tracking and precise estimation method for deviation between GNSS phase systems | |
| CN106646538A (en) | Single-difference filtering-based deformation monitoring GNSS (global navigation satellite system) signal multi-path correction method | |
| CN109459778A (en) | Code pseudorange based on robust variance component estimation/Doppler combines speed-measuring method and its application | |
| US20100265133A1 (en) | Method for suppressing multipath errors in a satellite navigation receiver | |
| CN111103600B (en) | A GPS/BDS multipath real-time suppression method based on normalization of single-frequency signal-to-noise ratio | |
| CN108196267B (en) | An uninterrupted time transfer method based on GNSS CP technology | |
| CN105738926A (en) | Method for calibrating phase inter-frequency bias between GLONASS system receiving machines | |
| CN109597105A (en) | A kind of GPS/GLONASS tight integration localization method for taking deviation between carrier system into account | |
| CN117706902B (en) | Clock stability evaluation method, device and medium | |
| CN114167457A (en) | GNSS system time deviation monitoring and evaluating method | |
| CN113671551A (en) | RTK positioning resolving method | |
| CN114879239B (en) | Regional three-frequency integer clock error estimation method for enhancing instantaneous PPP fixed solution | |
| CN106443729B (en) | A kind of adaptive pseudoranges of GNSS-phase power is than determining method | |
| CN109143289B (en) | GNSS single-station displacement monitoring method | |
| CN103760582B (en) | A kind of optimization method blocking satellite double-difference observation structure under environment | |
| CN118011434B (en) | A signal quality evaluation method for a positioning terminal with built-in antenna | |
| CN118465803B (en) | Real-time precise positioning method and system based on the characteristics of smartphone GNSS observation values | |
| CN111948682A (en) | Pseudo-range and carrier phase random model construction method based on BDS three-frequency combination | |
| CN109143286A (en) | A kind of satellite navigation locating method for taking non-model errors into account |
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |