## Abstract

This Paper describes the tuning of PID (proportional integral and derivative) by using a technique which is generally known as Particle Swarm Optimization. The technique reloves around the iterative method in terms of approach. In Reverse Osmosis System the method is used to adjust the pH level of the water in the system. The Particle Swarm Optimization is used to tune the parameters of PID controller. The purpose is to achieve an optimal solution by reaching to a stable and controlled system. The result can only be achieved by changing the variables of proportional integral and derivative controller.

## INTRODUCTION

he availability of clean water during the time is a major challenge. There are many ways to rely on water purification methods but as per some research these methods are not so efficient to provide drinkable water to a large number of population around the globe [1]. Major portion of water around ninety percent or even more is saltwater from oceans. Now desalination is a way to purify the saltwater and extract the excessive amount of salt in order to make it drinkable but the demand is still need to be met to keep up with the growing population. [3]

The Reverse Osmosis system operates on the principle of osmosis in which the particles in water pass through a membrane.[1] The system consists of a sediment filter, pre-carbon and a post carbon filter. The sediment filter is to reduce the number of particles in water. The pre-carbon filter consists of a granular surface and its function is to absorbe and removes chlorine in water.

The Post-carbon filters then removes the odour from the processed water and re-direct the water flow to water tank which is a storage unit. The concept of removing the salt in water comes from a membrane. A semi-permeable membrane which consist of thin layers and let the water flow pass through it, while rejecting a larger number molecules in water and dissolved salts. [4]

Throughout, researchers and scientists have tried and developed many ways to purify saltwater. Either through various filtration methods like distillation, deionization, carbon blocking, Ultra Violet filtration and Reverse Osmosis membrane method.

Controlling the pH level of water and Total dissolved solids (TDS) in water are one of the focused area where researchers try to eliminate the challenges and make the quality of water up to the standards. The pH level can be regulated by controlling the parameters of PID controller which is approached through particle swamp optimization method

## MODELING OF REVERSE OSMOSISSYSTEM

The water sample with the flux range, pressure, conductivity and pH level is taken and by using the transfer function we make some equations. Where flux and conductivity are taken as output and pressure and pH are taken as input. [7]

Save your time!

We can take care of your essay

- Proper editing and formatting
- Free revision, title page, and bibliography
- Flexible prices and money-back guarantee

Place Order
As in Reverse osmosis system if we change one variable it changes all other parameters along with it. By regulating pressure the flux of membrane change due to water flow capacity.

We have taken a range of variables in the table below. These variables can be regulated and their values are taken as samples. By changing one parameter we will find out the new behaviors in the system opposed to the previous values.

## TOTAL DISSOLVED SOLIDS AND MEMBRANE PERFORMENCE

Decaying of membrane can alter pH level. The membrane is mainly consist of layers and once the layers start to wear out it loose its effectively and alter the performance level which ultimately effect the pH level. Higher the pH rate more efficient the system would have to be in order to regulate it on an acceptable level. It contributes to underlying reason, which is the total dissolved solids (TDS level) more the dissolved particles are, greater the damage is done to the membrane.

## PARTICLE SWAM OPTIMIZATION

The particle swam optimization is an iterative method to approach a problem. The technique is relatively advanced [7]. It solves the problem by having number of solutions at the same time. The optimization technique works on the principle of iteration in a repetitive manner. After multiple iterations the probability of error reduces and gives an optimal solution. The particle is associated with position and velocity in terms of approaching a problem. The best value is achieved after updating the position and velocity of particle more than once. The movement of particles is guided according to the best search space in the algorithm. The selection principle is to move the particles and choose on the optimal value in comparison to all other solutions. In comparison with the other methods the particle swam optimization has better approach due the search and optimization.

The basic algorithm is then, for example the jth particle represented as xj = (xj1, xj2 , xj3 ,xj4…….xjn) in the g dimensional space. All the values of x are the numbers of particles in the search space. Then we begin the particle with the uniform position in the given search space. pj = xj . Then if (f(pj)≤f(xj)) once reached after iteration then it will update the new position relative to its velocity

Each particle has a track record in the search space according to its own unique dimension. Then it finds the best available position after searching repetitively. We can name the found value with p optimal. The best previous value of the jth particle is stored and it can be more than one value, not to neglect that each value has its unique position due to its optimal approach towards the given problem. So the best previous value can be represented as p-bestj=(pbestj1 pbestj2 pbestj3 …… pbestjg). Now comparing the finest of each particle with the global optimal position called as g-best is passed in the population as the current global optimal position g-best. In a similar manner with regard to position we have to consider the velocity as well. The velocity of the jth particle change is represented by vj=( vj1 ,vj2 ,vj3vjg ). Now it will keep changing the position and the search will keep on till the algorithm run out of iterations [7]:

## PROPOSED SCHEME OF PID TUNING

As mentioned previously that the particle swam optimization is the optimal method in term of finding the best possible value. Now we relate it to PID controller. PSO is used to tune the PID parameters (Kp, Ki, Kd The principle is defined as to accelerate number of particles in the search space and then regulate the parameters of PID controller. The PID parameters are adjusted in a specific way that they keep moving the particles and alter the position iteratively and ultimately reach to optimization.

## To start with the PSO certain parameters are to be selected for the optimum results. Initially the values of the population size and iteration are taken as 100. And velocity constant c1 and c2 are taken as 2. Fitness function for PSO is performance criteria which is Integral of Time multiplied by absolute Error (ITAE) is taken. CONCLUSION

As we have seen that conductivity is changed as we change the PH. Now PH depends on the total dissolved solids (TDS) which can be controlled once the water flows through the system. Conductivity can be controlled with PH levels and similarly flux can be controlled with the water pressure and damage control of the membrane. The particle swam optimization gives better results in terms of optimization but not to forget, that many losses and invisible conditions in real system have not been taken.

## REFERENCES

- Jimo Park, Goeun Kim, Jinsung Kim, Sanggun Na and HoonHeo” Simulation of reverse osmosis plant using RCGA based PID controller” ICROS-SICE International Joint Conference August 18-21,2009J.
- Gambier, A., A. Wellenreuther, and E. Badreddin. “Optimal control of a reverse osmosis desalination plant using multi-objective optimization.” Proceedings of the 2006 IEEE Conference on Control Applications, 1368- 1373, Munich, October 4-6,2006
- ImadAlatiqi*, HishamEttouney, Hisham El-Dessouky, “Process control in water desalination industry: an overview”, Desalination 126 (1999)15–32
- Abderrahim Abbas, “model Predictive control of a reverse osmosis Desalination unit”, Desalination 194(2006)268–280
- C. Riverol1 and V. Pilipovik2 ,“ Mathematical Modeling of Perfect Decoupled Control Systemand Its Application: A Reverse Osmosis Desalination Industrial-Scale Unit” , Journal of Automated Methods & Management in Chemistry, (2005), no. 2,50–54
- Jin-Sung Kim, Jin-Hwan Kim, Ji-Mo Park, Sung-Man Park, Won-Yong Choe, and HoonHeo “Auto Tuning PID Controller based on Improved Genetic Algorithm for Reverse Osmosis Plant” International Journal of Electrical and Computer Engineering 3:82008
- Natwar S. Rathore, Neha Kundariya, Anirudha Narain “ PID Controller Tuning in Reverse Osmosis System based on Particle Swarm Optimization” Electrical Engineering Department, MNNIT Allahabad