Lihui Bai

Simulation Modeling and Analysis of a Multi-resource Medical Clinic

Clinic of University of Louisville is a relatively new medical clinic which attempts to stem the epidemic of childhood obesity. The main problem addressed by this research is the no show rate (nearly 50%) of the clinic. There are two goals in this project. One is to increase the staff utilization; the other is to decrease the waiting time. We study two potential methods to solve this problem. One involves using multiple resources for every visit; the other involves overbooking the patients. One simulation model is an overbooking model in which the interarrival times are controlled for each type of patients. By increasing arrival rate of patients, the waiting time, the total number of served patients and the utilization of staff are increased. We need to trade off in order to choose the best arrival rate for the clinic. The second model involves using multiple resources for every visit. We also change the interarrival time for patients in order to estimate the best values for these inputs.

A simulation study for residential electricity user behavior under dynamic variable pricing with demand charge

ABSTRACT Attempting to increase energy efficiency and improve system load factors in an electricity distribution system, Demand Response (DR) has long been proposed and implemented as a form of load management. Various pricing structures incentivizing consumers to shift energy consumption from on-peak to off-peak periods are evident in this field. Most DR methods currently used in practice belong to static variable pricing (e.g., Time of Use, Critical Peak Pricing) and the impact of such tariffs has been well established. However, dynamic variable pricing in general is less studied and much less practiced in the field, due to the lack of understanding of consumer behavior in response to price uncertainty. In this article, we study a novel dynamic variable pricing scheme that uses the coincident demand charge to reduce load consumption during peak events. We employ a multi-attribute utility function and model predictive control to simulate consumer behavior of utility maximization in home energy consumption. We use a conditional Markov chain to model and predict the system peak. Effects of the proposed residential electricity rate based on coincident demand charge are compared with other pricing schemes through simulation validated with real-world residential load profiles. Finally, we extend the simulations to study the impact of integrating renewable solar production in a DR program.

Modelling and solution of a large-scale vehicle routing problem at GE appliances & lighting

We consider a special case of the vehicle routing problem where not only each customer has specified delivery time window, but each route has limited time duration. We propose a solution algorithm using network reduction techniques and simulated annealing meta-heuristic. The objective is twofold: minimising the travel time and minimising the total number of vehicles required. The time-window constraint ensures delivery without delay, thus, a potentially higher level of customer satisfaction. The algorithm has helped the transportation planning team at General Electric Appliances & Lighting to significantly reduce the number of required trucks in two real cases, while its performance on randomly generated cases is also efficient when compared to properly selected benchmarking algorithms.

Using agent based simulation and model predictive control to study energy consumption behavior under dynamic pricing

In the interest of increasing energy efficiency and avoiding higher generation costs during peak periods, utility companies adopt various demand response (DR) methods to achieve load leveling or peak reduction. DR techniques influence consumer behavior via incentives and cause them to shift peak loads to off-peak periods. In this paper we study the energy consumption behavior of residents in response to a variable real-time pricing function. We consider thermostatic loads, specifically air conditioning, as the primary load and apply the model predictive control (MPC) method to study the behavior of consumers who make consumption decisions based on a trade-off between energy cost and thermal comfort. An agent-based simulation is used to model a population where each household is an agent embedded with the MPC algorithm. Each household is associated with a multi-attribute utility function, and is uniquely defined via the use of stochastic parameters in the utility function.