"Nlfuzzy" is a NetLogo extension which acts as a a simple middleware between NetLogo and jFuzzyLogic library. It allows you to define fuzzy variables and rules for NetLogo agents (turtles, patches) using Fuzzy Control Language (see full FCL specification for more information).

Precompiled version can be downloaded here.

If you want to build nlfuzzy yourself, then your computer has to meet the following requirements:

• installed NetLogo 5.1 (5.0.x should work too) or at least NetLogo.jar library from the installation package,
• installed Java Development Kit (JDK) of at least version 6 (currently it is better to download version 8 which is backward compatible; see downloads at Oracle),
• installed Apache Maven.

Many Java libraries are available via public Maven repositories but in some cases, however, this is not true. This holds for both NetLogo and jFuzzyLogic which must be downloaded manually and installed as local Maven dependencies. Thanks to this, Maven will be able to find all the necessary dependencies defined in pom.xml file.

In case you have Windows Vista/7/8, open a command line window (cmd or powershell), change your working directory (command cd) to the downloaded and unpacked source code directory (i.e. the directory where file pom.xml is located) and type:

mvn install:install-file -Dfile="C:/Program Files (x86)/NetLogo 5.1.0/NetLogo.jar" -DgroupId=org.nlogo -DartifactId=netlogo -Dversion=5.1.0 -Dpackaging=jar

Parameter -Dfile must contain a correct path to NetLogo.jar according to where you installed NetLogo.

Download the core version of jFuzzyLogic from the project site and from nlfuzzy's project directory run:

mvn install:install-file -Dfile="C:/Users/JaneDoe/Downloads/jFuzzyLogic_core.jar" -DgroupId=net.sourceforge -DartifactId=jFuzzyLogic -Dversion=3.0 -Dpackaging=jar

Again, path to the respective jar file must be correct.

Finally, you can compile and package the project:

mvn clean dependency:copy-dependencies package

If you have built nlfuzzy from sources then copy the nlfuzzy.jar (located in the ./target directory) and jFuzzyLogic.jar (located in the ./target/dependency directory) files to your NetLogo installation directory's extensions/nlfuzzy subdirectory (extensions is already there but nlfuzzy has yet to be created).

In case you have downloaded the compiled version, just copy the nlfuzzy directory to the extensions directory.

If you have a running NetLogo instance then you must exit the program and run it again.

Before attaching any rules or evaluating any variables, an nlfuzzy:init procedure must be called (typically in the "setup" function - see the example below).

to setup
  nlfuzzy-init
end

Nlfuzzy is able to operate with multiple rule sets. Each rule set is defined using a single FCL file. Rule file is attached to an agent via function nlfuzzy:use-fcl path_to_a_file.fcl.

ask turtles [
  nlfuzzy-use-fcl "my-fuzzy-rules.fcl"
]

Input variables (defined in your FCL file) are set via function *nlfuzzy:set-value "fuzzy-variable-name" "a-value-to-be-set".

ask turtles [
  nlfuzzy:set-value "some_fuzzy_variable" "a_value_to_be_assigned"
]

This function evaluates defined rules for passed output variable name. E.g.:

ask turtles [
  set speed nlfuzzy:eval "speed"
]

A skeleton of a nlfuzzy-based program may look like this:

extensions [ nlfuzzy ]

to setup
  nlfuzzy:init
  ask agents [
    nlfuzzy:use-fcl "my_fcl_rules.fcl"
end

to go
  ...
  ask turtles [
    nlfuzzy:set-value "some_fuzzy_variable" "a_value_to_be_assigned"
  ]
  ...
  ask turtles [
    set some-variable (nlfuzzy-eval "output_variable")
  ]
]

A simple working demo showing how to use nlfuzzy can be found in the demo directory.