An Adaptive Prediction Filter Algorithm for Optimizing Servo System Performance,
Abstract
Several approaches have been taken by the Instrumentation Directorate over the past few years to improve the servo performance of various types of optical tracking telescopes. These efforts have concentrated on servo systems having sample rates between 20 and 200 per second where the original approaches centered around using analog techniques to implement lag-lead compensation networks comprised of R-C networks and operational amplifiers. Around 1970, efforts were successful in utilizing a special purpose mini-computer to implement a purely digital servo compensation as called for from a sampled-data analysis. In 1974, a compensation method was devised that used a non-linear gain relationship of the servo error signal to control a high performance mirror mount with a system sample rate of 200 per second. A mini-computer was used to implement the compensation and excellent performance results were obtained. This approach suggested that different methods could be used for servo control than had previously been tried and that such techniques could provide satisfactory servo performance. The experience gained from these earlier approaches coupled with the technique to be described have resulted in the development of an Adaptive Prediction Filter (APF) which allows a simple procedure to be followed in compensating a servo system and yields an optimum performance. (Author)
Document Details
- Document Type
- Technical Report
- Publication Date
- Jan 01, 1976
- Accession Number
- ADA025980
Entities
People
- Lester M. Bradley