A dynamic simulation model designed to optimize farmer profits and minimize gin processing and energy consumption provides the basis for computer selection of gin machinery sequences at the Stoneville Laboratory. The dynamic model, reported in 1982 by Anthony, Wesley and Brown, utilizes the market price structure for cotton and the machinery performance characteristics for each gin machine to determine the optimum machinery sequence. Application of the model decisions in the recently renovated small-scale research facility (microgin) at Stoneville was achieved by installing special routing valves to bypass or select any combination of four seed-cotton cleaners, two multi-path driers, and three lint cleaners as directed by a computer. Three dimensional decision matrices that contain machinery decisions based on measurements of moisture, color and foreign matter are used in the control system. Measurements are made at three stations in the gin system: 1) Feed control, 2) feed hopper above the extractor-feeder, and 3) battery condenser. Stations 1 and 2 evaluate seed cotton whereas station 3 evaluates lint.

Initial measurements of moisture with an infrared moisture meter, and color and foreign matter with a High Volume Instrument (HVI) color/trash meter are used to determine the number of shelves of drying required as well as the number of seed cotton cleaners. Drier options differ from those in a conventional ginning system. In our system 0, 12, 18, or 24 shelves in each of two stages of drying with static air temperatures are available. In conventional systems two stages of 24 shelves are used and the temperature is adjustable. Measurements at station 1 are used to select the gin machinery processing sequence, and to predict values for moisture, color and foreign matter at station 2. Predicted values are compared with the actual values when the cotton arrives at station 2. Differences are then used to index the values at station 1 to compensate for differences in cotton cleanability and machine performance.

Measurements at station 2 are also used to select the number of lint cleaners. Subsequent measurements of color and foreign matter at station 3 are compared to those predicted from station 2 and used to index entry values into the machinery decision matrices at station 2. Color and foreign matter measurements at station 3 establish the final values for the ginned lint before baling.

Analog voltages from the moisture meters are digitized by an OPTO-22(1) board. HVI digital signals are input via an additional serial port board. Output signals control the cotton flow through the equipment by switching pneumatic directional valves. These output signals are transmitted via the OPTO-22 board through solid-state relays. A PC-compatible microcomputer and C language form the basis for the control system.

The gin process control system at the Stoneville Lab has been operated successfully, Additional improvements are underway before large-scale testing and subsequent incorporation into a conventional ginning system are pursued.