981677 The Fuzzy Inference System Translator (FIST) and
Contents
1. If temperature is NORMAL then coldwater is NORMAL 1 0 If temperature is HOT then coldwater is FAST 1 0 If humidity is MOIST then humidifier is OFF 1 0 If humidity is DRY then humidifier is ON 1 0 If humidity is WET then humidifier is OFF 1 0 If humidity is MOIST then humidifier is OFF 1 0 If temp is NORMAL and humidity is WET then heater is ON 1 0 If temp is NORMAL and humidity is WET then coldwater is FAST 1 0 ALS Fuzzy Inference System Logic Toolbox off low hans A Heater fuzzy slow m fast Temperature logic A 12 rules a N Cold Water off os Y Humidity Humidifier Figure 2 The ALS fuzzy system showing the fuzzified inputs and crisp outputs Upon development or modification of the fuzzy inference system in Matlab the relevant information about it must be converted into C code and attached to the main C pro gram fuzzy c through header h files Our Matlab func tion ged m creates these header files automatically This m file is available for downloading at our website Dur ing compilation the program fuzzy c is translated into a format compatible with the Motorola 68HC11 micropro cessor We have also made available our source code files written in Control C of the Mosaic Industries QED Program Development System see Figure 3 Software Development Environment MATLAB2. QED Fuzzy Contr ol C Logic System Toolbox Tools programming download Control C Code Software re 68HC11 ofthe Code Standalone compilation Fuzzy Logic Controller Figure 3 Software development environment showing the relationship between MATLAB and the QED program development system The program fuzzy c is a multitasking application provid ing control functions as well as permanent interactive communication with the user see Figure 4 In the con trol mode it reads input analog signals from ALS sensors and invokes the fuzzy logic algorithm to compute the desired control signals for heater CHX and humidifier It also makes non fuzzy decisions for other control signals photoperiod control nutrient delivery and CO injection Current values of all input and output signals are dis played on the LCD screen Structure of the FIST Software M ain Program fuzzy c Specific Program celss c Analog Signal Reading W riting Functions Calibration Functions Interactive M essages General Keyboard and Display Functions Non fuzzy Logic M odified Matlab Fuzzy Logic Function and the Description of Fuzzy System Pra include celss c Figure 4 Structure of the FIST software showing the relationship between the main program fuzzy c and the ALS specific program celss c The continuous loop during normal operation of the AL
3. S controller can be interrupted by the user at any moment by depressing the next button on the keypad The ALS controller then switches to its Interactive Calibration Mode In the interactive mode the user can calibrate the sensors and change the nominal operating points of the controlled environmental parameters We separated all the specific codes of the ALS controller such as mes sages and the calculation of non fuzzy control signals into the program celss c to make the software more adaptable to other ALS applications This program is written in the C control language of the QED Mosaic pro gram development system FUZZY LOGIC CONTROLLER The compiled hexadec imal code is then downloaded via an RS232 serial cable to the 128K memory chip on the QED product design board The following 0 5 V sensor voltage signals are connected to the input pins of the QED analog digital convertor RTD temperature humidity water level pres sure and CO sensors The 0 5 V actuator signals for the heater chilled water valve position humidifier nutri ent pump lights and CO injector are supplied by the analog output latch of the QED digital analog converter This stand alone controller keeps the ALS temperature and humidity near the desired operational levels Figure 5 shows the relative humidity and temperature within the ALS along with the ambient laboratory temperature for a 24 hr light dark cycle In addition to temperature and humidity co
4. SAE TECHNICAL PAPER SERIES 981677 The Fuzzy Inference System Translator FIST and Micro Controller Regulation of Plant Growth Chamber Temperature and Humidity Bill Taylor Elena Leyderman James Vredenburg Andr s Estrada and Janell Kueffer New Mexico Highlands University Anthony Maestas Hughes Aircraft The Engineering Society 28th International nferen Ga Se ie i Hi Mobility st te nat onal Conference Land Sea Air and Space on Environmental Systems INTERNATIONAL Danvers Massachusetts July 13 16 1998 e o aM a aa EE EE 400 Commonwealth Drive Warrendale PA 15096 0001 U S A Tel 724 776 4841 Fax 724 776 5760 The appearance of this ISSN code at the bottom of this page indicates SAE s consent that copies of the paper may be made for personal or internal use of specific clients This consent is given on the condition however that the copier pay a 7 00 per article copy fee through the Copyright Clearance Center Inc Operations Center 222 Rosewood Drive Danvers MA 01923 for copying beyond that permitted by Sec tions 107 or 108 of the U S Copyright Law This consent does not extend to other kinds of copying such as copying for general distribution for advertising or promotional purposes for creating new collective works or for resale SAE routinely stocks printed papers for a period of three years following date of publication Direct your orders to SAE Customer Sales and Satisfaction Department Quantity reprin
5. e Motorola 68HC11 micro controller Software Approach The Matlab Fuzzy Logic Toolbox was used for fuzzy controller program development 3 We specified fuzzy membership functions for the temper ature and relative humidity signals Next a set of weighted inference rules was developed from our own experience and working knowledge of the ALS environ mental system This experience came from several culti var grow outs first using set point control and then with the fuzzy micro controller hardware regimen The first step was to develop a mathematical model of plant growth chamber dynamics based on the depen dence of temperature on heat load and water vapor pres sure on temperature Our experimental data from the ALS system subjected to a variety of step inputs in tem perature and humidity showed a first order exponential with no overshoot response We therefore used a simple first order dynamic model simulated on Simulink for rapid prototyping of the fuzzy inference system To describe the temperature at the current time step 7 in the absence of a plant canopy we used the following dif ference equation T T kH ko Tin T Al Eq 1 where H is the heat load and T is the laboratory tem perature For vapor pressure P we used the static rela tionship b nh a Eq 2 k where a and b are constants determined experimentally The rationale behind this approach is the belief we have not yet veri
6. fied this that fuzzy ALS control will be insensi tive to parameter changes nonlinearities and higher order effects The simple model described above implemented in Sim ulink allowed us to test our ideas for fuzzy membership functions and fuzzy rules interactively with the Matlab Fuzzy Logic Toolbox The prototype fuzzy logic controller then was tested in the laboratory using Simulink and the Matlab Real Time Workshop to accept sensor inputs and provide actuator signals to the ALS control elements At this point we had a working fuzzy inference system one that required our computer to be connected directly to the ALS The next step in the development process was the design of the fuzzy inference system translator FIST This is the software tool that makes it possible to implement fuzzy controllers developed with the Matlab Fuzzy Logic Toolbox on stand alone microprocessors In this opera tion all relevant fuzzy inference system information is extracted from the Matlab fis matrix Both the modified Matlab fuzzy inference algorithm and the extracted fuzzy inference system parameters are processed so that the resulting machine code generated by the FIST software will run on the Motorola 68HC11 micro controller In this technical paper we describe the FIST program and the resulting ALS system controller see Figure 1 Fuzzy Logic Controller Outputs Inputs Fuzzy Fuzzy Heater _ QED Box with Analog and Digital Board
7. gSystems ABOUT THE MAIN AUTHOR Bill Taylor received his doctorate from the University of California Davis in 1989 and currently is an Associate Professor of Engineering at New Mexico Highlands Uni versity in Las Vegas New Mexico
8. l Both nutrient delivery and CO concentration are regulated by simple set point control System cooling and de humidification functions are pro vided by condensing heat exchangers CHX located Bill Taylor Elena Leyderman James Vredenburg Andr s Estrada and Janell Kueffer New Mexico Highlands University Anthony Maestas Hughes Aircraft beneath each of the cabin enclosures Each CHX con nects to a dedicated recirculating cooler with coolant flow metered by a voltage controlled linear valve Airflow through the CHX is established but not controlled by a pair of blowers Any required heating is supplied by an electric heater while a household humidifier contributes water vapor on demand Data Acquisition System Hewlet Packard VEE and the XVI data acquisition mainframe with a digitizing voltmeter and 32 multiplexed input channels form the basis of the data acquisition system Thermistor temperature read ings are collected at several locations within each PGC cabin enclosure and CHX output stream Also output voltages from the humidity sensor and CO sensor in each testbed are monitored along with the output of a pressure transducer that detects the presence of stand ing water in the rooting media Also monitored by the data acquisition system are the actuator voltages of the ALS environmental control system FUZZY CONTROLLER Originally all environmental control functions for our ALS were regulated by set poin
9. l sys tems such as advanced life support systems plant growth chambers and greenhouses The FIST program gives an adaptable alternative to classical control of advanced life support and environmental systems The resulting ALS control system appears to be robust and insensitive to changes in dynamics as the cultivar matures Further system tests are planned to test these hypotheses 24 hr Test of ALS Controller 80 ra 60 4 g o gt e 5 50 BE E 40 Lights On sa Lights Off On ao E 2 305 ALS temp os icine E a 20 4 ambient temp 10 4 0 11 12 13 15 1617181921 2223 0 2 3 4 6 7 8 1011 Time hr Figure 5 Twenty four hour test of fuzzy logic controller over one 24 hr light dark cycle of the ALS testbed The relative humidity upper plot stabilizes near its nominal value of 65 after two hours Interior ALS temperature is maintained near its nominal value of 25 C regardless of heat load or ambient temperature ACKNOWLEDGMENTS The authors gratefully acknowledge the generous sup port of Kennedy Space Center through grant number NAG10 0161 REFERENCES 1 Taylor B and G Zrili Closed loop testing of a controlled environmental life support system CELSS for a space based cultivar abstract Annals of Biomedical Engineer ing vol 21 suppl 1 p 24 1993 2 Taylor B and G Zrili A fuzzy logic controller for a con trolled ecological life
10. ntrol the ALS controller provides desired pho toperiod control supervises appropriate nutrient delivery and allows the operator to adjust the richness of the car bon dioxide environment Table II lists typical nominal val ues of the ALS environmental variables Table Il Nominal values of ALS environmental variables Temperature 25C Relative Humidity 65 CO Concentration 1000 ppm SUMMARY AND CONCLUSIONS FUZZY LOGIC CONTROL There are two modes to the operation of the fuzzy logic controller the Interactive Cal ibration Mode and the Control Mode In the Interactive Calibration Mode there is one stage for the calibration of input sensors and a second stage for setting the thresh old levels of plant growth chamber environmental vari ables Also the user can set the controller clock and the ON OFF times for the desired photoperiod All calibrated data are battery backed Control Mode In the Control Mode the fuzzy controller is running and provides all actuator signal voltages The Control Mode of the ALS controller enables the display of input and or output values With the controller keypad and LCD the user can switch at anytime between modes and switch between functions within each mode Figure 6 A Flexible Approach to ALS Control The Fuzzy Infer ence System Translator FIST is a valuable tool in the development of fuzzy logic software for use on stand alone micro controllers for controlled environmenta
11. s a Motorola 68H C11 Microprocessor y Temperature Pa aad MEAory GIID 4 7 za x old W ater Humidity A p Humidifier L 7 _ Non fuzzy Non fuzzy Water Level OED co CO Injector 2 Lights Figure 1 General scheme of the ALS system controller with both fuzzy and non fuzzy inputs and outputs FUZZY CONTROLLER SOFTWARE DEVELOPMENT FUZZY INFERENCE SYSTEM The first step in fuzzy controller software development is to fuzzify the envi ronmental signals of interest using the Matlab Fuzzy Logic Toolbox For example temperature ranges may be designated as cool normal or hot Obviously the desired nominal operating temperature should occur somewhere in the middle of the normal range see Figure 2 Simi larly humidity levels are divided into possibly overlapping ranges of dry moist and wet Next a set of fuzzy rules are specified such as if temperature is normal and humidity is dry then coldwater is slow where the flow rate of the chilled water stream may be slow medium or fast A complete set of fuzzy rules currently used in the ALS control system are presented in Table I Table I Fuzzy rules for an ALS control system including the weight assigned to each rule If temperature is COOL then heater is ON 1 0 If temperature is NORMAL then heater is LOW 0 1 If temperature is HOT then heater is OFF 1 0 If temperature is COOL then coldwater is SLOW 1 0
12. support system IN M Jamshidi C Nguyen R Lumia and J Yuh eds ntelligent Automation and Soft Computing vol 1 pp 613 618 1994 3 Jang J S R and N Gulley Fuzzy Logic Toolbox User s Guide The MathWorks Inc Natic MA 1995 Functional Block Diagram of the FIST Software Multitasking Application eo P N a gt a f d gt I I Interactive ControllerTask Keypad Scanning Task user interrupt alw ays active gather sensor data waitfor key pressor ealcucteioutntcaltage invoke appropriate actions STOPPED RUNNING control output devices run controller stop controller NEXT key NEXT key calibrate sensor switch display display input input level set input or output or output data Figure 6 Functional block diagram of the FIST software for both the controller running mode and the interactive keypad mode SOFTWARE TOOLS AVAILABLE ON THE INTER NET The interested reader will find an interactive controller simulation in Java on the world wide web using our set of fuzzy membership functions and fuzzy rules see Figure 2 The complete source code for FIST which can serve as a bridge between fuzzy inference system development on Matlab and implementation on the QED Product Development System and a complete User s Manual for the Fuzzy Logic Controller also are available at our web site http vyne nmhu edu Livin
13. t control first with a computer workstation running HP VEE and then with a dedicated Motorola microprocessor Using simple set point control we were not able to achieve our desired temperature and humidity levels simultaneously In our attempt to overcome some of the problems inherent in the control of coupled temperature and humidity dynamics we investigated the potential use of fuzzy controllers Hardware Approach For our first attempt at fuzzy logic control we programmed and tested the NeuroLogix sin gle chip fuzzy micro controller Environmental signals such as temperature humidity and water level were selected for inputs to the fuzzy micro controller Then the chip was programmed to compute the degrees of belong ing to fuzzy sets assigned to each channel These were evaluated in parallel within the fuzzy micro controller generating outputs to control the heater blowers and the nutrient delivery pump This work has been reported elsewhere 2 Two growth cycles of a chile cultivar were completed using this fuzzy logic hardware approach We found the fuzzy micro controller approach to be lim ited by the number of channels and the number of rules that could be processed simultaneously Also the speed of this chip 1 10 Mhz far exceeded the requirements of our ALS control system Accordingly we decided to investigate the potential of fuzzy logic software imple mentation on an inexpensive microprocessor in our case th
14. t rates can be obtained from the Customer Sales and Satisfaction Department To request permission to reprint a technical paper or permission to use copyrighted SAE publications in other works contact the SAE Publications Group GLOBAL MOBILITY DATABASE All SAE papers standards and selected books are abstracted and indexed in the Global Mobility Database No part of this publication may be reproduced in any form in an electronic retrieval system or otherwise without the prior written permission of the publisher ISSN 0148 7191 Copyright 1998 Society of Automotive Engineers Inc Positions and opinions advanced in this paper are those of the author s and not necessarily those of SAE The author is solely responsible for the content of the paper A process is available by which discussions will be printed with the paper if it is published in SAE Transactions For permission to publish this paper in full or in part contact the SAE Publications Group Persons wishing to submit papers to be considered for presentation or publication through SAE should send the manuscript or a 300 word abstract of a proposed manuscript to Secretary Engineering Meetings Board SAE Printed in USA 981677 The Fuzzy Inference System Translator FIST and Micro Controller Regulation of Plant Growth Chamber Temperature and Humidity Copyright 1998 Society of Automotive Engineers Inc ABSTRACT The Fuzzy Inference System Translator FIST is a
15. tool in the realization of standalone fully programmable fuzzy logic micro controllers for the regulation of advanced life support system temperature and humidity subsystems Analog input signals may include chamber temperature relative humidity CO concentration and nutrient level Analog output signals can be for example heater voltage and condensing heat exchanger cold water valve volt age nutrient pump actuator voltage and grow lamp actu ator voltage Features of the micro controller described include keypad entry of sensor calibration data and online modification of the photo period temperature humidity and CO levels during full system operation All system inputs and outputs can be selected for read out on a liquid crystal display LCD INTRODUCTION LIFE SUPPORT SYSTEM TESTBED The advanced life support ALS system testbed in operation at New Mexico Highlands University consists of twin sealed environmentally controlled cabin enclosures 1 The vol ume of each clear Plexiglas enclosure is 1 0 m contain ing a Phototron plant growth chamber PGC complete with flourescent lights and electrical supply A removable hatch provides access to the PGC for cleaning pruning and harvesting Plant nutrients are delivered by a positive displacement pump to a sphagnum moss rooting medium Atmospheric CO levels are maintained with a bottled supply system to replace photosynthetic uptake by the plants Environmental Contro
Download Pdf Manuals
Related Search