Francisco Jose Moreno-Velo

FPGA Implementation of Embedded Fuzzy Controllers for Robotic Applications

Fuzzy-logic-based inference techniques provide efficient solutions for control problems in classical and emerging applications. However, the lack of specific design tools and systematic approaches for hardware implementation of complex fuzzy controllers limits the applicability of these techniques in modern microelectronics products. This paper discusses a design strategy that eases the implementation of embedded fuzzy controllers as systems on programmable chips. The development of the controllers is carried out by means of a reconfigurable platform based on field-programmable gate arrays. This platform combines specific hardware to implement fuzzy inference modules with a general-purpose processor, thus allowing the realization of hybrid hardware/software solutions. As happens to the components of the processing system, the specific fuzzy elements are conceived as configurable intellectual property modules in order to accelerate the controller design cycle. The design methodology and tool chain presented in this paper have been applied to the realization of a control system for solving the navigation tasks of an autonomous vehicle.

Automatic design of fuzzy controllers for car-like autonomous robots

This paper describes the design and implementation of a fuzzy control system for a car-like autonomous vehicle. The problem addressed is the diagonal parking in a constrained space, a typical problem in motion control of nonholonomic robots. The architecture proposed for the fuzzy controller is a hierarchical scheme which combines seven modules working in series and in parallel. The rules of each module employ the adequate fuzzy operators for its task (making a decision or generating a smoothly varying control output), and they have been obtained from heuristic knowledge and numerical data (with geometric information) depending on the module requirements (some of them are constrained to provide paths of near-minimal lengths). The computer-aided design tools of the environment Xfuzzy 3.0 (developed by some of the authors) have been employed to automate the different design stages: 1) translation of heuristic knowledge into fuzzy rules; 2) extraction of fuzzy rules from numerical data and their tuning to give paths of near-minimal lengths; 3) offline verification of the control system behavior; and 4) its synthesis to be implemented in a true robot and be verified on line. Real experiments with the autonomous vehicle ROMEO 4R (designed and built at the Escuela Superior de Ingenieros, University of Seville, Seville, Spain) demonstrate the efficiency of the described controller and of the methodology followed in its design.

Design of Embedded DSP-Based Fuzzy Controllers for Autonomous Mobile Robots

Fuzzy controllers are used in many applications because of their rapid design by translating heuristic knowledge, robustness against perturbations, and smoothness in the control action. However, they require parallel processing and special operators (such as fuzzification or defuzzification) which are not available in standard digital signal processors (DSPs), thus complicating their direct implementation. This paper describes an efficient design methodology that allows starting with any kind of fuzzy controller and subsequently transforming it until a system suitable for easy DSP implementation is obtained. Such methodology is greatly aided by the design environment Xfuzzy 3. The parking problem of an autonomous robot is described to illustrate the steps of this methodology. Real experiments with the autonomous robot ROMEO 4R demonstrate efficiency of the designed fuzzy controller embedded into a stand-alone card based on a fixed-point DSP from Texas Instruments.

Using Xfuzzy Environment for the Whole Design of Fuzzy Systems

Since 1992, Xfuzzy environment has been improving to ease the design of fuzzy systems. The current version, Xfuzzy 3, which is entirely programmed in Java, includes a wide set of new featured tools that allow automating the whole design process of a fuzzy logic based system: from its description (in the XFL3 language) to its synthesis in C, C++ or Java (to be included in software projects) or in VHDL (for hardware projects). The new features of the current version have been exploited in different application areas such as autonomous robot navigation and image processing.

CAD Tools for Hardware Implementation of Embedded Fuzzy Systems on FPGAs

This paper describes two computer-aided design (CAD) tools for automatic synthesis of fuzzy logic-based inference systems. The tools share a common architecture for efficient hardware implementation of fuzzy modules, but are based on two different design strategies. One of them is focused on the generation of standard VHDL code, which can be later implemented on a reconfigurable device [field-programmable gate array (FPGA)] or as an application-specific integrated circuit (ASIC). The other one uses the Matlab/Simulink environment and tools for development of digital signal processing (DSP) systems on Xilinxs FPGAs. Both tools are included in the last version of Xfuzzy, which is a specific environment for designing complex fuzzy systems, and they provide interfaces to commercial VHDL synthesis and verification tools, as well as to conventional FPGA development environments. As demonstrated by the included design example, the proposed development strategies speed up the stages of description, synthesis, and functional verification of embedded fuzzy inference systems.

Tuning complex fuzzy systems by supervised learning algorithms

Tuning a fuzzy system to meet a given set of input/output patterns is usually a difficult task that involves many parameters. This paper presents an study of different approaches that can be applied to perform this tuning process automatically, and describes a CAD tool, named xfsl, which allows applying a wide set of these approaches: (a) a large number of supervised learning algorithms; (b) different processes to simplify the learned system; (c) tuning only specific parameters of the system; (d) the ability to tune hierarchical fuzzy systems, systems with continuous output (like fuzzy controller) as well as with categorical output (like fuzzy classifiers), and even systems that employ user-defined fuzzy functions; and, finally, (e) the ability to employ this tuning within the design flow of a fuzzy system, because xfsl is integrated into the fuzzy system development environment Xfuzzy 3.0.

NORFREA: An algorithm for non redundant fuzzy rule extraction

This contribution presents a new algorithm (NORFREA) to select fuzzy rules from a grid partition of the input domain. Besides using an efficiency measure for the rules, this algorithm employs an heuristic technique to reduce the influence of the initial grid structure. Different benchmarks of classification problems are included to illustrate the advantages of this algorithm.

XFSML: An XML-based modeling language for fuzzy systems

This paper presents a new modeling language for fuzzy systems called XFSML. It is an XML-based language and it is proposed as a starting point for the definition of a standard modeling language in the fuzzy community. The main features of the language are its high expressiveness and its independence from specific platforms, tools or programming languages.

Embedded fuzzy controllers on standard DSPs

Fuzzy controllers are used in many applications because of their rapid design by translating heuristic knowledge, robustness against perturbations, and smoothness in the control action. However, they require parallel processing and special operators (such as fuzzification or defuzzification) which are not available at standard DSPs, thus making inefficient its direct implementation. This paper describes a design methodology which allows starting with any kind of fuzzy controller and subsequently transforming it until obtaining a system suitable for DSP implementation. Such methodology is aided by Xfuzzy 3, a design environment developed by some of the authors. The parking problem of an autonomous robot is described to illustrate the steps of this methodology. Experimental results show the efficiency of the designed fuzzy controller embedded into a stand-alone card based on a fixed-point DSP from Texas Instruments.

FuzzyCN2: An algorithm for extracting fuzzy classification rule lists

Most of the algorithms for extracting fuzzy classification rules generate conjunctive antecedents that use all the attributes of the system. Using this kind of antecedents, the number of rules grows exponentially in terms of the number of attributes of the system. This paper presents a new algorithm, FuzzyCN2, for extracting conjunctive fuzzy classification rules. This algorithm is a fuzzy version of the well known CN2 algorithm and produces an ordered list of fuzzy rules. FuzzyCN2 generates antecedents that may not include all the attributes of the system. These antecedents may cover a wide number of instances and, so, the number of extracted rules is smaller. The algorithm introduces the use of linguistic hedges as part of the selectors, thus producing more compact rules and reducing the number of generated rules.