Accuracy design of the instrument

1. Determination of instrument accuracy indicators

The basis for determining the accuracy of instruments comes from national standards, ministerial standards, tasks assigned by superiors, and requirements put forward by the user unit, but in the final analysis it is determined by the use requirements.

The accuracy index of measuring instruments is generally taken as 1/10 to 1/3 of the total use requirements (i.e. the total measurement result accuracy). For instruments with general accuracy requirements, it is 1/10, and for high-precision instruments, it is 1/3. This is a rough judgment based on experience and is not very reasonable, but in situations where environmental errors and operating errors are difficult to estimate accurately, this empirical data is still used.

2. Determination of instrument system error

The value of the system error calculated and synthesized according to the above method should generally not be greater than 1/3 of the total instrument error. If the system error exceeds this range, it means that the initial data is unreasonable, and the design should be modified, or the process requirements should be improved, or compensation measures should be considered.

3. Error distribution - the relationship between component error and total error

(1) Equal tolerance method

The equal tolerance method is to give the same tolerance value to each error source, that is,

Where Δqi is the original error of each error source.

Although the equal tolerance method is simple, it is not reasonable because the manufacturing difficulty of each part of the instrument is different, and the error sources are

The transfer coefficients are also different.

(2) Equal Impact Method

The equal influence method is to make the contribution of the error source to the total error of the instrument equal. In the formula, dy/dq is the transfer coefficient of each original error Δqi.

The equal influence method takes into account the influence of the transfer coefficient and is more reasonable than the equal tolerance method. However, since the process conditions are not taken into account, the Δqi calculated by the equal influence method may be too high or too low. Therefore, Δ￠qi needs to be adjusted according to the difficulty of the actual process.

(3) Trial method

The trial and error method is to give an error allocation scheme based on the designer's experience, and then perform verification and adjustment. It is also possible to conduct experiments on certain key links to provide data. For example, for a certain aiming method, you can first conduct an experiment to find out the value of the aiming error without waiting for allocation.

(4) Optimization design method

Some people have proposed to take the total error of a single measurement of the instrument as the objective function, the errors of each link in the instrument measurement chain and the construction parameters as design variables, and the range of variation of each parameter and the range of tolerance as constraints to form a constrained minimization problem, which is optimization design.

The optimization design method is undoubtedly an advanced method. Today, when electronic computers are widely used, it is feasible to use computer-aided design.