Sabo Miya Hassan

Adopting EWMA Filter on a Fast Sampling Wired Link Contention in WirelessHART Control System

The performance of a wireless industrial process plant is affected by wired and wireless links reliability. While recent related works focus mainly on wireless links reliability, this paper focuses on wired links reliability between the controller and the wireless highway addressable remote transducer (WirelessHART) node. In control application for a process plant with fast sampling, when the wired link delay is too short, message transmissions to the node could occur concurrently resulting in packet collision and packet dropout, thus leading to local contention. This can cause process plant with critical control application to be unstable. Therefore, this paper proposes an effective design of exponentially weighted moving average filter to mitigate the contentions effect and improve the process plants performance. In order to make evident and validate the existence of local contention, an experiment is carried out using a WirelessHART device. From the experiment, the controller delay is 0.1475 s and the estimated packet dropout is 12.29%. Simulation results with the designed filter show a significant improvement in process performance in terms of percentage of overshoot compared with performance of the process plant using traditionally optimized PID controller.

Simulation of WirelessHART networked control system with packet dropout

Industrial communication standard like WirelessHART enables wireless control in process plants. This paper presents a study on wired and wireless networked control system. It emphasizes on using packet dropout array for Markov chain simulation of packet dropout process, a key parameter affecting plants performance, in a closed-loop WirelessHART networked control system. Additionally, Bernoulli simulation of the process is covered for comparison purpose. Overshoot was found to be 633.3 % for Bernoulli model, 33.9 % for Markov model and even through it is the major drawback of the system, WirelessHART networked control system is still suitable for control applications when non-time-critical plant has response time longer than sampling period thousands times even with 50 %, 40 % dropout probabilities for Bernoulli, Markov models.

Setpoint weighted wirelesshart networked control of process plant

The use of conventional PID controllers has proven to be inadequate in delayed and uncertain networks such as WirelessHART network. This is because the PIDs produce oscillatory and unstable response when large gains are used. Furthermore, the response become sluggish when very small gains are used. Because of complexity, the use of deadtime compensators and other model based controllers in the industries is also difficult. In this work, the application of setpoint weighting control strategy for WirelessHART environment is presented. The method permits the use of high stable gain and provides for good load disturbance regulation. Comparison of the performance of the proposed method to both PID and Smith predictor structures showed that it is better in terms of response speed and overshoot. It is also robust to load disturbance. Hence it provides a two degree of freedom control.

Analyzing WirelessHART High-level Data Link Control frame payload

WirelessHART has become an interesting technology for enabling wireless monitoring and control applications in industrial process plants. However, only few tools are available for studying and investigating WirelessHART network and messages such as Wi-HTest. This paper introduces an approach to analyze and study WirelessHART High-level Data Link Control frame payload. It utilizes Eltimas Serial Port Monitor and MATLAB® software to provide a dedicated tool for reading, analyzing and generating the frame payload to interface with a WirelessHART development module like XDM2510H. The paper also points out and proves the existence of several security threats with WirelessHART protocol such as traffic analysis, eavesdropping and tampering.

Internal model control for industrial wireless plant subject to measured round trip delays

This paper presents a study of Internal Model Control (IMC) scheme to improve control performance of industrial wireless networked control system subject to measured round trip delays. As compared to the auto-tuned optimized PID controller, IMC controller helps to significantly reduce percentage of overshoot for a first order with dead-time process plant. The paper also highlights the differences in control performance of the process plant with lumped network delay and separate network delays.

Predictive PI controller for wireless control system with variable network delay and disturbance

As control using wireless technology is receiving increased attention, its often associated with network delays and uncertainties. The use of PIDs proved inadequate to handle these challenges while model based controllers are complex. As a compromise, this paper examines the use of Predictive PI (PPI) controller in a wireless networked control environment characterized by both network delay and process deadtime. As against deadtime compensators, the PPI allows for model mismatch while still maintaining the simplicity of PIDs in terms of tunable parameters. Comparison of the simulation results for the thermal chamber while considering input disturbance and measurement noise showed that the PPI controller produce improved performance in terms of systems overshoot, rise time and integral absolute error in all cases. The PPI is still better than the other two controllers in terms of settling time for most cases.

Effect of network induced delays on WirelessHART control system

Delay has been one of the critical factors affecting control performance of a wireless networked control system. It leads to the instability of the system. Network induced delay awareness is important for investigating the delay effects on the system. There are several direct and indirect ways to measure delays in a network. This paper concentrates on a direct delay measurement approach between WirelessHART controller and its plant. Upstream and downstream delays are measured by calculating the difference at timestamps between two consecutive received messages at the controller and at the plant respectively. The measured roundtrip delay is used to study delay effects on the control systems performance.

Towards the development of WirelessHART adaptor for process control applications

Recent advances in wireless technologies have prompted the applications of these cheap technologies to industrial monitoring and control application. However, this application is hindered by lack of adequate infrastructure such as the wireless adaptors. This led to the development of adaptors by organization such as Emerson Process Management, AwiaTech Corporation, etc. The developed adaptors by these organizations are expensive and not easily interoperable with devices from other manufacturers. To overcome this problem, this work proposes development of a low cost and interoperable adaptor. The adaptor will be developed around a wireless mote and a microcontroller. Preliminary results show that it is possible to perform two level control with the adaptor as shown from the experiment.

Optimal PID control of pH neutralization plant

pH control plays an important role in a wide range of industrial applications for waste management. The high non-linearity in pH neutralization processes and the uncertainty of the plant dynamics are key challenges that cannot be effectively addressed using PID controllers tuned with conventional methods. Some of the available and quite often tuning methods used for designing PID controllers are Cohen-Coon, Ziegler-Nichols (ZN), and Tyreus-Luyben. To address these problems, particle swarm optimization (PSO) is employed to tune the PID controller for a pH neutralization plant. The PSO algorithm is one of the most powerful methods for resolving non-smooth global optimization problems. In this paper, the mathematical model of pH process is developed taking into consideration of the overall plants dynamics. The PSO based PID controller is then implemented and evaluated by comparing its performance with that of ZN-PID.

Neural network architecture selection for efficient prediction model of gas metering system

This paper presents a comparative study and analysis of different neural network architectures of which one will be recommended towards adoption for developing a prediction model for gas metering system. Thus, the focus of this paper is to select the most suitable neural network architecture for gas metering system prediction model. A few neural networks architecture are modeled and simulated; Radial basis Function (RBF), Multilayer Perceptron (MLP), Elman Network, Generalized Regression Neural Networks (GRNN) and Elman Neural Network. In order to select the best architecture, the performance of the various networks considered are compared. From the results obtained, the network architecture that results in the best performance is the RBF network structure. Hence recommended for adoption for the design.