J.M.G. Sá da Costa

Soft Computing Approaches to Fault Diagnosis for Dynamic Systems

Recent approaches to fault detection and isolation for dynamic systems using methods of integrating quantitative and qualitative model information, based upon soft computing (SC) methods are surveyed and studied in some detail. SC methods are considered an important extension to the quantitative model-based approach for residual generation in fault detection and isolation (FDI). When quantitative models are not readily available, a correctly trained neural network (NN) can be used as a non-linear dynamic model of the system. The paper describes some powerful NN methods, taking into account the dynamic as well as non-linear system behaviour. Sometimes, further insight is required as to the explicit behaviour of the model-involved and it is then that fuzzy and even neurofuzzy methods come to their own in data-driven FDI applications. The paper also discusses the use of evolutionary programming tools for observer and NN design. The paper provides many powerful examples of the use of SC methods for achieving good detection and isolation of faults in the presence of uncertain plant behaviour, together with their practical value for fault diagnosis of real process systems.

An architecture for fault detection and isolation based on fuzzy methods

Model-based fault detection and isolation (FDI) is an approach with increasing attention in the academic and industrial fields, due to economical and safety related matters. In FDI, the discrepancies between system outputs and model outputs are called residuals, and are used to detect and isolate faults. This paper proposes a model-based architecture for fault detection and isolation based on fuzzy methods. Fuzzy modeling is used to derive nonlinear models for the process running in normal operation and for each fault. When a fault occurs, fault detection is performed using the residuals. Then, the faulty fuzzy models are used to isolate a fault. The FDI architecture proposed in this paper uses a fuzzy decision making approach to isolate faults, which is based on the analysis of the residuals. Fuzzy decision factors are derived to isolate faults. An industrial valve simulator is used to obtain several abrupt and incipient faults, which are some of the possible faults in the real system. The proposed fuzzy FDI architecture was able to detect and isolate the simulated abrupt and incipient faults.

Distributed supply chain management using ant colony optimization

Successful supply chain management requires a cooperative integration between all the partners in the network. At the operational level, the partners individual behavior should be optimal and therefore their activities have to be planned using sophisticated optimization tools. However, these tools should take into account the planning of the remaining partners, through the exchange of information, in order to allow some kind of cooperation between the elements of the chain. This paper introduces a new supply chain management technique, based on modeling a generic supply chain with suppliers, logistics and distributers, as a distributed optimization problem. The different operational activities are solved by the optimization meta-heuristic called ant colony optimization, which allows the exchange of information between different optimization problems by means of a pheromone matrix. The simulation results show that the new methodology is more efficient than a simple decentralized methodology for different instances of a supply chain.

Optimization problems in multivariable fuzzy predictive control

Abstract The application of model predictive control (MPC) to complex, nonlinear processes results in a non-convex optimization problem for computing the optimal control actions. This optimization problem can be solved by discrete search techniques such as the branch-and-bound method (B&B), which has been successfully applied to MPC. However, the discretization induced by B&B introduces a tradeoff between the number of discrete actions and the performance. This paper proposes a solution for non-convex optimization problems in multiple-input multiple-output (MIMO) systems. Fuzzy predictive filters, which are represented as an adaptive set of control actions multiplied by gain factors, are extended for MIMO systems. This solution keeps the number of necessary alternatives low and increases the performance. The proposed MPC method using fuzzy predictive filters is applied to the control of a gantry crane. Simulation results show the advantages of the proposed method.

Distributed optimisation of a logistic system and its suppliers using ant colonies

This paper introduces a new multi-agent approach for collaborative management of logistic and supply systems based on the ant colony optimisation (ACO) meta-heuristic. The logistic system and its suppliers can be modelled as partners of a supply chain. The management methodology is defined as a set of distributed scheduling problems that exchange information during the optimisation process. Each problem is solved by an ant colony agent that uses the pheromone matrix as the communication platform. A simulation example shows that the proposed coordination mechanism improves the supply-chain performance compared to a traditional management approach, where both problems are considered separately.

Fuzzy predictive algorithms applied to real-time force control

Abstract This paper proposes combining a classical impedance controller with a fuzzy predictive algorithm. This algorithm calculates the optimal virtual trajectory that is given to the impedance controller. This control strategy allows for the inclusion of a non-rigid environment, represented by a nonlinear model, in the control design in a straightforward way. Thus, improving the global force control performance. In order to reduce the oscillations of the optimized reference position, a fuzzy scaling machine is included in the force control strategy. The performance of the force control scheme is illustrated for an experimental two-degree-of-freedom robot. A real-time implementation of the fuzzy predictive algorithm revealed better performance in terms of force control than the classical force control algorithms.

Fuzzy active noise modeling and control

Abstract The design of active noise control (ANC) has been developed in the last two decades based on linear identification and control tools. However, acoustic processes present nonlinearities coming both from the characteristics of the actuator and from the nature of the process. Recent research has emphasized the importance of nonlinear model-based controllers, which increase the performance of several types of systems. From the different nonlinear techniques, fuzzy modeling is one of the most utilized. Direct and inverse multivariable fuzzy models can be identified directly from data using fuzzy clustering. Inverse models can then be applied directly as controllers, which can be included in an active noise control scheme. This paper proposes the use of fuzzy techniques in ANC. The performance of the proposed control schemes is compared to classical finite impulse response ANC in an experimental setup. The proposed fuzzy control scheme outperforms classical active noise controllers.

A multi-agent approach for supply chain management using ant colony optimization

Distributed systems like supply chains can be efficiently managed by multi-agent approaches. However, the control of each sub-system in a supply chain is a complex optimization problem where optimal performance can be achieved using meta-heuristics. This paper presents a new methodology for supply chain management, a distributed optimization using ant colonies, where the concepts of agents and ant colony optimization are merged, using the pheromone matrix as communication platform. This paper formalizes this methodology, using as an example a supply chain with logistic, supplying and distribution sub-systems.

Optimization of Logistic Processes in Supply-Chains Using Meta-heuristics

This paper addresses the optimization of logistic processes in supply-chains using meta-heuristics: genetic algorithms and ant colony optimization. The dynamic assignment of components to orders and choosing the solution that is able to deliver more orders at the correct date, is a scheduling problem that classical scheduling methods can not cope with. However, the implementation of meta-heuristics is done only after a positive assessment of the performance’s expectation provided by the fitness-distance correlation analysis. Both meta-heuristics are then applied to a simulation example that describes a general logistic process. The performance is similar for both methods, but the ant colony optimization method provides more information at the expenses of computational costs.

Pattern recognition in power-system security

Abstract Power system security assessment is of great importance in power-system operation. To date, security evaluation has consisted of occasionally performing load flows and transient stability studies by computer simulation of current system status or by applying conventional direct methods. This type of security evaluation is very time-consuming and of no practical use in large modern power systems. The paper describes the general philosophy behind an overall security system based on pattern-recognition theory. Problems related to the design and field implementation are discussed.