Maizura Mokhtar

Sensor Failure Detection, Identification, and Accommodation Using Fully Connected Cascade Neural Network

Modern control systems rely heavily on their sensors for reliable operation. Failure of a sensor could destabilize the system, which could have serious consequences to the systems operations. Therefore, there is a need to detect and accommodate such failures, particularly if the system in question is of a safety critical application. In this paper, a sensor failure detection, identification, and accommodation (SFDIA) scheme is presented. This scheme is based on the fully connected cascade (FCC) neural network (NN) architecture. The NN is trained using the neuron by neuron learning algorithm. This NN architecture is chosen because of its efficiency in terms of the number of neurons and the number of inputs required to solve a problem. The SFDIA scheme considers failures in pitch, roll, and yaw rate gyro sensors of an aircraft. A total of 105 experiments were conducted; out of which, only one went undetected. The SFDIA scheme presented here is efficient, compact, and computationally less expensive, in comparison to schemes using, for example, the popular multilayer perceptron NN. These benefits are inherited from the FCC NN architecture.

Microgrid development for properties

Microgeneration and/or on-site generation from renewable sources, which are generally intermittent, have been pushing the way forward to introduce intelligence to Building Management System (BMS), so as to integrate local energy generation and storage with the public grid and form an intelligent microgrid. This paper aims to develop a property microgrid for the Samuel Lindow Building of the University of Central Lancashire at its Westlakes Campus, West Cumbria, UK. This system covers energy (both electricity and heat) generation from on-site renewable sources and gas, energy storage and energy demand management. Through optimum operation control of the microgrid whilst minimising electrical power transfer from the main grid, this system will enable the property to maximise the benefit of the renewable sources, to improve energy efficiency, to reduce the energy bill, to minimise the impact of the renewable generation on the main grid, and to reduce greenhouse gas emission.

An ARTMAP-incorporated multi-agent system for building intelligent heat management

This paper presents an ARTMAP-incorporated multi-agent system (MAS) for building heat management, which aims to maintain the desired space temperature defined by the building occupants (thermal comfort management) and improve energy efficiency by intelligently controlling the energy flow and usage in the building (building energy control). Existing MAS typically uses rule-based approaches to describe the behaviours and the processes of its agents, and the rules are fixed. The incorporation of artificial neural network (ANN) techniques to the agents can provide for the required online learning and adaptation capabilities. A three-layer MAS is proposed for building heat management and ARTMAP is incorporated into the agents so as to the facilitate online learning and adaptation capabilities. Simulation results demonstrate that ARTMAP-incorporated MAS provides better (automated) energy control and thermal comfort management for a building environment in comparison to its existing rule-based MAS approach.

Comparing the online learning capabilities of Gaussian ARTMAP and Fuzzy ARTMAP for building energy management systems

Recently, there has been a growing interest in the application of Fuzzy ARTMAP for use in building energy management systems or EMS. However, a number of papers have indicated that there are important weaknesses to the Fuzzy ARTMAP approach, such as sensitivity to noisy data and category proliferation. Gaussian ARTMAP was developed to help overcome these weaknesses, raising the question of whether Gaussian ARTMAP could be a more effective approach for building energy management systems? This paper aims to answer this question. In particular, our results show that Gaussian ARTMAP not only has the capability to address the weaknesses of Fuzzy ARTMAP but, by doing this, provides better and more efficient EMS controls with online learning capabilities.

Aircraft sensor estimation for fault tolerant flight control system using fully connected cascade neural network

Flight control systems that are tolerant to failures can increase the endurance of an aircraft in case of a failure. The two major types of failure are sensor and actuator failures. This paper focuses on the failure of the gyro sensors in an aircraft. The neuron by neuron (NBN) learning algorithm, which is an improved version of the Levenberg-Marquardt (LM) algorithm, is combined with the fully connected cascade (FCC) neural network architecture to estimate an aircrafts sensor measurements. Compared to other neural networks and learning algorithms, this combination can produce good sensor estimates with relatively few neurons. The estimators are developed and evaluated using flight data collected from the X-Plane flight simulator. The developed sensor estimators can replicate a sensors measurements with as little as 2 neurons. The results reflect the combined power of the NBN algorithm and the FCC neural network architecture.

A multi-objective planning framework for optimal integration of distributed generations

This paper presents an evolutionary algorithm for analyzing the best mix of distributed generations (DG) in a distribution network. The multi-objective optimization aims at minimizing the total cost of real power generation, line losses and CO2 emissions, and maximizing the benefits from the DG over a 20 years planning horizon. The method assesses the fault current constraint imposed on the distribution network by the existing and new DG in order not to violate the short circuit capacity of existing switchgear. The analysis utilizes one of the highly regarded evolutionary algorithm, the Strength Pareto Evolutionary Algorithm 2 (SPEA2) for multi-objective optimization and MATPOWER for solving the optimal power flow problems.

Increasing endurance of an autonomous robot using an Immune-Inspired framework

This paper describes the implementation of an online immune-inspired framework to help increase endurance of an autonomous robot. Endurance is defined as the ability of the robot to exert itself for a long period of time. The immune-inspired framework provides such capability by monitoring the behavior of the robot to ensure continuous and safe behavior. The immune-inspired framework combines innate and adaptive immune inspired algorithms. Innate uses a dendritic cell based innate immune algorithm, and adaptive uses an instance based B-cell approach. Results presented in this paper shows that when the robot is implemented with the immune-inspired framework, health and survivability of a robot is improved, therefore increasing its endurance.

Power Profiling and Inherent Lag Prediction of a Wind Power Generating System for Its Integration to an Energy Storage System

A key challenge within the power sector is to address the issue of intermittency. It is the unavailability of energy at all times in order to meet the demand requirements. Intermittency is responsible for reducing the efficiency of the national infrastructure and can compromise energy security. Increasing use of renewable energy can cause the increasing intermittency. This is an important issue that needs to be dealt with. Predictive mechanisms based on historical data have been used previously to try and address energy security with renewables. However, the effectiveness of the predictive mechanisms are low. Going forward, energy storage systems will play a key role in securing the energy supply provided by renewables. Efficient use of energy storage relies on information about the generator system that it is coupled with. This paper aims to show that despite the inherent characteristics of renewable energy generation, the nature of mechanical generation of renewable systems can be equated and modelled. The model can provide the information required for energy storage coupling. The model equates the inherent lag using the torque values of the generator, as well as the generators velocity. The model is part of a larger framework that predicts the output power profile of the renewables, using an Artificial Neural Network (ANN). The predictive information can further improve the performance of the coupled energy storage system and address intermittency.

Understanding the regulation of predatory and anti-prey behaviours for an artificial organism

An organisms behaviour can be categorised as being either predatory or anti-prey. Predatory behaviours are behaviours that try to improve the life of an organism. Anti-prey behaviours are those that attempt to prevent death. Regulation between these two opposing behaviours is necessary to ensure survivability--and gene regulatory networks and metabolic networks are the mechanisms that provide this regulation. We know that such regulatory behaviour is encoded in an organisms genes. The question is, how is it encoded? The understanding of this encoding can help with the development of an artificial organism, for example an autonomous robotic system; whereby the robot will have the ability to autonomously regulate the switching between the opposing behaviours using this encoded mechanism, in order to ensure its sustainable and continuous system operations. This paper aims to look into the properties of an artificial bio-chemical network consisting of a genetic regulatory network and a metabolic network that can provide these capabilities.

Adaptive and Online Health Monitoring System for Autonomous Aircraft

ight, especially for an Unmanned Aerial System (UAS). Good situation awareness can be achieved by incorporating an Adaptive Health Monitoring System (AHMS) to the aircraft. The AHMS monitors the ight outcome or ight behaviours of the aircraft based on its external environmental conditions and the behaviour of its internal systems. The AHMS does this by associating a health value to the aircraft’s behaviour based on the progression of its sensory values produced by the aircraft’s modules, components and/or subsystems. The AHMS indicates erroneous ight behaviour when a deviation to this health information is produced. This will be useful for a UAS because the pilot is taken out of the control loop and is unaware of how the environment and/or faults are aecting the behaviour of the aircraft. The autonomous pilot can use this health information to help produce safer and securer ight behaviour or fault tolerance to the aircraft. This allows the aircraft to y safely in whatever the environmental conditions. This health information can also be used to help increase the endurance of the aircraft. This paper describes how the AHMS performs its capabilities.