Study On The Quality Loss Function And Calibration Of Measurement System

Modern quality management emphasizes on results, while the results depend on process. The basis of product quality improvement is to objectively evaluate the quality level, which involves two problems:first, whether the method to evaluate the quality level is comprehensive, accurate and effective; Second, whether the data used to evaluate is reliable, in other words, the quality of the data from the measurement system is good or not. We can make use of some correlation theory of quality loss function to solve the first problem, while to the second, first of all, we must ensure a high quality of the measurement system, then, we ought to design the reasonable calibration cycle and calibration methods for the measurement system, what's more, we should describe the quality of measuring level through the measurement error loss function. Measuring quality loss function and the calibration methods of measuring system are the important aspect of the taguchi method, which have extremely extensive application in characterizing the quality of measuring level, determining the best calibration cycle of measuring system and the calibration method of the measurement system, evaluating and improving product quality, designing of quality control system, and so on. Thus, the study to the quality loss function, the measurement error loss function, the feedback control systems, the calibration system loss function and the calibration method of non-linear measurement system has an important value not only on theory but also on application.In previous studies, the author found that there were still some shortages in measuring quality loss function and related theory itself, which caused widespread controversy in academia since the taguchi method was born, and to some extent, it reduced the enthusiasm of industrial enterprise to use this method. Based on the study of the related theories, we found problems and put forward the calibration method of the loss function and non-linear measuring system after some corresponding improvement. This paper is organized as follows:(1) Improvement of quality loss function. At present, Taguchi quality loss function which without exception only by a quadratic term. It's unadvisable for the smaller the better and larger the better situations. This paper comes up with the improved quality loss function of smaller the better and larger the better characteristic. Then, this paper studies the form of dynamic quality characteristic loss function.(2) Forms of measurement error loss function in dynamic measurements. Taguchi measurement quality engineering introduces the measurement error loss function in static measurement. However, in the practice, the measured object is always changed dynamically. This paper comes up with the measurement loss function in dynamic measurements.(3) Improvement of feedback control system loss function. Taguchi designed the feedback control system loss function based on considering measuring cost, setup cost and loss of out-of-control product. But Taguchi's the feedback control system loss function still has rooms for improvement. This paper comes up with the improved Taguchi feedback control system loss function.(4) Designing of the control chart based on improved feedback control system loss function. This paper transforms some parameters which in improved feedback control system loss function so as to design the control chart based on feedback control system loss function which make the least economic loss. Then studies the control chart how to apply in measurement system stability and bias analysis.(5) Improvement of calibration system loss function. On the basis of finding out the problems which exist in Taguchi calibration system loss function, the author comes up with the improved calibration system loss function.(6) Study on the calibration cycle and calibration method of the linear measurement system. This paper takes the CMM using in measuring dimensions as case. On the basis of analyzing the current calibration cycle and calibration method of the CMM using in measuring dimensions, this paper put forward the optimum calibration cycle and calibration method of the CMM using in measuring dimensions based on Taguchi measurement quality engineering.(7) Study on the calibration method of non-linear measurement system. Most measurement systems are non-linear, especially the measurement systems which contains sensors and is used to measure various flow. For non-linear measurement system, it will increase system error artificially if processed by linear relation. This paper studies the non-linear calibration method of measurement system by referencing the theory in Taguchi measurement quality engineering.