粗差也被称为过失误差或异常误差。测量者的主观因素或测试条件突然变化导致

粗差也被称为过失误差或异常误差。测量者的主观因素或测试条件突然变化导致的明显测量结果不一致的误差。例如，仪表操作不当，是由于读数错误、记录和计算错误、测试系统突然故障和环境条件等不注意因素造成的误差。粗差观测值被称为不好的值或异常值，用一定的方法判别除去，不影响测量结果。<br>在水平测量平差计算中，通常使用概率统计方法来消除观测值之间的不一致值，这是一系列偶然误差的观测值。对于与偶然误差同时存在的各种系统误差，为了在观测过程中消除或减弱对观测结果的影响，为了检查在闭路差是否超过限度时观测值是否存在粗糙差在平差计算中没有考虑系统误差和粗糙度。<br>关于粗差，其位置和大小没有事先调查，粗差只影响极为个别的观测数据。不能适用消除补偿和影响的自我检查平差法。近年来，我们开发了两种基本的粗差处理方法：识别和调节。识别是一种勘探，探测后删除。调整是指采用抗干扰性强的估计方法，控制粗差对平差结果的影响。本节主要讨论去除粗差的方法。

Roughness is also known as error of error or abnormal error. The error of the obvious measurement results due to subjective factors of the surveyor or sudden changes in test conditions. For example, improper instrument operation is due to errors caused by factors such as incorrect readings, incorrect recording and calculation, sudden failure of the test system, and environmental conditions. Rough observations are called bad values or outliers, and they are determined to be removed by a certain method, without affecting the measurement results.<br>In horizontal measurement of flatness calculations, probability statistics are often used to eliminate inconsistencies between observations, which are observations of a series of accidental errors. For various system errors that exist at the same time as accidental errors, in order to eliminate or reduce the influence on the observation results during the observation process, in order to check whether the observations have rough differences when the closed-circuit difference exceeds the limit, the system error and roughness are not taken into account in the flatness calculation.<br>With regard to coarseness, its location and size were not pre-investigated, and coarseness affected only very individual observations. The self-checking balance method to eliminate compensation and effects cannot be applied. In recent years, we have developed two basic rough handling methods: identification and adjustment. Identification is an exploration that is removed after detection. Adjustment refers to the use of strong anti-jamming estimation method to control the effect of roughness on the results of the spread. This section focuses on ways to remove roughness.

Gross error is also called gross error or abnormal error. The measurement error caused by subjective factors or sudden changes of test conditions. For example, improper operation of instruments is caused by reading errors, recording and calculation errors, sudden failure of test system and environmental conditions. Gross error observation value is called bad value or abnormal value, which can be discriminated and removed by certain methods, and does not affect the measurement results.<br>In the calculation of horizontal survey adjustment, the probability and statistics method is usually used to eliminate the inconsistent values between observations, which are a series of observations with accidental errors. In order to eliminate or weaken the influence on the observation results in the observation process, and to check whether there is roughness in the observation value when the closed-circuit error exceeds the limit, the systematic error and roughness are not considered in the adjustment calculation.<br>As for the gross error, its location and size are not investigated in advance, and the gross error only affects very individual observation data. The self checking adjustment method of eliminating compensation and influence can not be applied. In recent years, we have developed two basic gross error processing methods: recognition and adjustment. Recognition is a kind of exploration, which is deleted after detection. Adjustment refers to the use of strong anti-interference estimation method to control the influence of gross errors on the adjustment results. This section mainly discusses how to remove gross errors.<br>