Figure 2 two-dimensional fuzzy PI controller structure type

The quantitative factor Ke, Kec will input variable from basic theory field into the corresponding fuzzy set theory field, scale factor Ku the fuzzy control algorithm is given control quantity switch to rotating joints control object can accept basic theory field. After analysis and a large number of simulation experiments, a quantization factor and scaling factor effect the performance of the system are as follows:
(1) Ke effect the performance of the system The greater the Ke, system control inert is smaller, faster rate rise; Ke is too large, system rise rate is too large, produce overshoot is too large, the regulating time lengthen, even produce oscillation and system unstable; Ke is too small, the system rise rate is too small, regulating inert change, at the same time also influence the system steady performance, steady state accuracy reduce.
(2) Kec will effect the performance of the system Kec is bigger, the system state changes of the inhibition ability increase, improves the system's stability; Kec is too large, system rise rate is too small, the system transition time extension; Kec is too small, the system rise rate increase may lead to system to produce big overshoot or oscillation.
(3) Ku effect the performance of the system Ku is equivalent to rotary union the system total magnification. Ku increases, the system response speed; Ku is too big, can cause the system rise rate is too large, which can lead to big overshoot or oscillation or divergent; Ku is too small, system prior to gain is small, the system rise rate is lesser, rapidity variation, steady state accuracy variation. 3, output scale factor self-tuning adaptive fuzzy control In this paper, an online scale factor self-adjusting fuzzy controller, as shown in figure 3 shows, it is in the main fuzzy controller based on, the introduction of a auxiliary fuzzy reasoning, in the control process according to fiber optic rotary joint the error and error change size and relationship, create a control volume to revise scale factor Ku, so as to improve the control performance of the controller. Scale factor Ku adjustment quantity alpha and error E and error change EC relationship can be expressed as: α (K) = f [E (K), EC (K)]
(1) Type of f for E and EC nonlinear function, the value of α totally depends on system transient state, and the controlled object model and has nothing to do. This kind of based on the scale factor of self adjustment method is actually a kind of independent of the model of the nonlinear variable gain controller. (1) the main fuzzy controller input/output a definition of the variable Main fuzzy controller with error E and error change EC as input to control the quantity change value U as output. Error, error change and output language value fuzzy subset for {NB, NM, NS, ZE, PS, PM, PB} seven state, region established.fuzzy for [1, 1].
(2) auxiliary fuzzy controller input/output a definition of the variable Auxiliary fuzzy controller with error E and error change EC as input to Ku correction quantity α as output. Revised scale factor: Ku = α × Ku. Error and error change language value fuzzy subset and main fuzzy controller, α language value fuzzy subset for {ZE, VS, S, SB, MB, B, VB}, its basic theory field range for .