Reinforcement Learning With Function Approximation for Traffic Signal Control

We propose, for the first time, a reinforcement learning (RL) algorithm with function approximation for traffic signal control. Our algorithm incorporates state-action features and is easily implementable in high-dimensional settings. Prior work, e. g., the work of Abdulhai et al., on the application of RL to traffic signal control requires full-state representations and cannot be implemented, even in moderate-sized road networks, because the computational complexity exponentially grows in the numbers of lanes and junctions. We tackle this problem of the curse of dimensionality by effectively using feature-based state representations that use a broad characterization of the level of congestion as low, medium, or high. One advantage of our algorithm is that, unlike prior work based on RL, it does not require precise information on queue lengths and elapsed times at each lane but instead works with the aforementioned described features. The number of features that our algorithm requires is linear to the number of signaled lanes, thereby leading to several orders of magnitude reduction in the computational complexity. We perform implementations of our algorithm on various settings and show performance comparisons with other algorithms in the literature, including the works of Abdulhai et al. and Cools et al., as well as the fixed-timing and the longest queue algorithms. For comparison, we also develop an RL algorithm that uses full-state representation and incorporates prioritization of traffic, unlike the work of Abdulhai et al. We observe that our algorithm outperforms all the other algorithms on all the road network settings that we consider.

Multi-agent Reinforcement Learning for Traffic Signal Control

Optimal control of traffic lights at junctions or traffic signal control (TSC) is essential for reducing the average delay experienced by the road users amidst the rapid increase in the usage of vehicles. In this paper, we formulate the TSC problem as a discounted cost Markov decision process (MDP) and apply multi-agent reinforcement learning (MARL) algorithms to obtain dynamic TSC policies. We model each traffic signal junction as an independent agent. An agent decides the signal duration of its phases in a round-robin (RR) manner using multi-agent Q-learning with either is an element of-greedy or UCB 3] based exploration strategies. It updates its Q-factors based on the cost feedback signal received from its neighbouring agents. This feedback signal can be easily constructed and is shown to be effective in minimizing the average delay of the vehicles in the network. We show through simulations over VISSIM that our algorithms perform significantly better than both the standard fixed signal timing (FST) algorithm and the saturation balancing (SAT) algorithm 15] over two real road networks.

Modeling urban traffic planning thro' strategic decisions- A case from Bangalore, India

This paper deals with reducing the waiting times of vehicles at the traffic junctions by synchronizing the traffic signals. Strategies are suggested for betterment of the situation at different time intervals of the day, thus ensuring smooth flow of traffic. The concept of single way systems are also analyzed. The situation is simulated in Witness 2003 Simulation package using various conventions. The average waiting times are reduced by providing an optimal combination for the traffic signal timer. Different signal times are provided for different times of the day, thereby further reducing the average waiting times at specific junctions/roads according to the experienced demands.

Performance of cellular CDMA with voice/data traffic with an SIR based admission control

We analyze the performance of an SIR based admission control strategy in cellular CDMA systems with both voice and data traffic. Most studies In the current literature to estimate CDMA system capacity with both voice and data traf-Bc do not take signal-tlFlnterference ratio (SIR) based admission control into account In this paper, we present an analytical approach to evaluate the outage probability for voice trafllc, the average system throughput and the mean delay for data traffic for a volce/data CDMA system which employs an SIR based admission controL We show that for a dataaniy system, an improvement of about 25% In both the Erlang capacity as well as the mean delay performance is achieved with an SIR based admission control as compared to code availability based admission control. For a mixed voice/data srtem with 10 Erlangs of voice traffic, the Lmprovement in the mean delay performance for data Is about 40%.Ah, for a mean delay of 50 ms with 10 Erlangs voice traffic, the data Erlang capacity improves by about 9%.

Reinforcement learning with average cost for adaptive control of traffic lights at intersections

We propose for the first time two reinforcement learning algorithms with function approximation for average cost adaptive control of traffic lights. One of these algorithms is a version of Q-learning with function approximation while the other is a policy gradient actor-critic algorithm that incorporates multi-timescale stochastic approximation. We show performance comparisons on various network settings of these algorithms with a range of fixed timing algorithms, as well as a Q-learning algorithm with full state representation that we also implement. We observe that whereas (as expected) on a two-junction corridor, the full state representation algorithm shows the best results, this algorithm is not implementable on larger road networks. The algorithm PG-AC-TLC that we propose is seen to show the best overall performance.

An instantaneous control model for flow control in an ATM network

This paper is concerned with the optimal flow control of an ATM switching element in a broadband-integrated services digital network. We model the switching element as a stochastic fluid flow system with a finite buffer, a constant output rate server, and a Gaussian process to characterize the input, which is a heterogeneous set of traffic sources. The fluid level should be maintained between two levels namely b1 and b2 with b1control on the ATM network is considered

Some aspects of multilevel load-frequency control of a power system

The application of multilevel control strategies for load-frequency control of interconnected power systems is assuming importance. A large multiarea power system may be viewed as an interconnection of several lower-order subsystems, with possible change of interconnection pattern during operation. The solution of the control problem involves the design of a set of local optimal controllers for the individual areas, in a completely decentralised environment, plus a global controller to provide the corrective signal to account for interconnection effects. A global controller, based on the least-square-error principle suggested by Siljak and Sundareshan, has been applied for the LFC problem. A more recent work utilises certain possible beneficial aspects of interconnection to permit more desirable system performances. The paper reports the application of the latter strategy to LFC of a two-area power system. The power-system model studied includes the effects of excitation system and governor controls. A comparison of the two strategies is also made.

Some aspects of multilevel load-frequency control of a power system

The application of multilevel control strategies for load-frequency control of interconnected power systems is assuming importance. A large multiarea power system may be viewed as an interconnection of several lower-order subsystems, with possible change of interconnection pattern during operation. The solution of the control problem involves the design of a set of local optimal controllers for the individual areas, in a completely decentralised environment, plus a global controller to provide the corrective signal to account for interconnection effects. A global controller, based on the least-square-error principle suggested by Siljak and Sundareshan, has been applied for the LFC problem. A more recent work utilises certain possible beneficial aspects of interconnection to permit more desirable system performances. The paper reports the application of the latter strategy to LFC of a two-area power system. The power-system model studied includes the effects of excitation system and governor controls. A comparison of the two strategies is also made.

Jointly optimal power control and routing for a single cell, dense, ad hoc wireless network

We consider a dense, ad hoc wireless network confined to a small region, such that direct communication is possible between any pair of nodes. The physical communication model is that a receiver decodes the signal from a single transmitter, while treating all other signals as interference. Data packets are sent between source-destination pairs by multihop relaying. We assume that nodes self-organise into a multihop network such that all hops are of length d meters, where d is a design parameter. There is a contention based multiaccess scheme, and it is assumed that every node always has data to send, either originated from it or a transit packet (saturation assumption). In this scenario, we seek to maximize a measure of the transport capacity of the network (measured in bit-meters per second) over power controls (in a fading environment) and over the hop distance d, subject to an average power constraint. We first argue that for a dense collection of nodes confined to a small region, single cell operation is efficient for single user decoding transceivers. Then, operating the dense ad hoc network (described above) as a single cell, we study the optimal hop length and power control that maximizes the transport capacity for a given network power constraint. More specifically, for a fading channel and for a fixed transmission time strategy (akin to the IEEE 802.11 TXOP), we find that there exists an intrinsic aggregate bit rate (Theta(opt) bits per second, depending on the contention mechanism and the channel fading characteristics) carried by the network, when operating at the optimal hop length and power control. The optimal transport capacity is of the form d(opt)((P) over bar (t)) x Theta(opt) with d(opt) scaling as (P) over bar (1/eta)(t), where (P) over bar (t) is the available time average transmit power and eta is the path loss exponent. Under certain conditions on the fading distribution, we then provide a simple characterisation of the optimal operating point.

Experimental Demonstration of H{infty} Control based Active Vibration Suppression in Composite Fin-tip of Aircraft using Optimally Placed Piezoelectric Patch Actuators

The goal of this study is the multi-mode structural vibration control in the composite fin-tip of an aircraft. Structural model of the composite fin-tip with surface bonded piezoelectric actuators is developed using the finite element method. The finite element model is updated experimentally to reflect the natural frequencies and mode shapes accurately. A model order reduction technique is employed for reducing the finite element structural matrices before developing the controller. Particle swarm based evolutionary optimization technique is used for optimal placement of piezoelectric patch actuators and accelerometer sensors to suppress vibration. H{infty} based active vibration controllers are designed directly in the discrete domain and implemented using dSpace® (DS-1005) electronic signal processing boards. Significant vibration suppression in the multiple bending modes of interest is experimentally demonstrated for sinusoidal and band limited white noise forcing functions.

Robust flight stability and control for micro air vehicles.

This chapter presents the real time validation of fixed order robust 112 controller designed for the lateral stabilisation of a micro air vehicle named Sarika2. Digital signal processor (DSP) based onboard computer named flight instrumentation controller (FIC) is designed to operate under automatic or manual mode. FIC gathers data from multitude of sensors and is capable of closed loop control to enable autonomous flight. Fixed order lateral H-2 controller designed with the features such as incorporation of level I flying qualities, gust alleviation and noise rejection is coded on to the FIC. Challenging real time hardware in loop simulation (HILS) is done with dSPACE1104 RTI/RTW. Responses obtained from the HILS are compared with those obtained from the offline simulation. Finally, flight trials are conducted to demonstrate the satisfactory performance of the closed loop system. The generic design methodology developed is applicable to all classes of Mini and Micro air vehicles.

A voltage-sensorless control method to balance the input currents of a three-wire boost rectifier under unbalanced input voltages condition

This paper proposes a control method that can balance the input currents of the three-phase three-wire boost rectifier under unbalanced input voltage condition. The control objective is to operate the rectifier in the high-power-factor mode under balanced input voltage condition but to give overriding priority to the current balance function in case of unbalance in the input voltage. The control structure has been divided into two major functional blocks. The inner loop current-mode controller implements resistor emulation to achieve high-power-factor operation on each of the two orthogonal axes of the stationary reference frame. The outer control loop performs magnitude scaling and phase-shifting operations on current of one of the axes to make it balanced with the current on the other axis. The coefficients of scaling and shifting functions are determined by two closed-loop prportional-integral (PI) controllers that impose the conditions of input current balance as PI references. The control algorithm is simple and high performing. It does not require input voltage sensing and transformation of the control variables into a rotating reference frame. The simulation results on a MATLAB-SIMULINK platform validate the proposed control strategy. In implementation Texas Instrument's digital signal processor TMS320F24OF is used as the digital controller. The control algorithm for high-power-factor operation is tested on a prototype boost rectifier under nominal and unbalanced input voltage conditions.

Comparative study of adaptive frequency hopping with power control to avoid WLAN interference in WPAN systems like Bluetooth

Presently Bluetooth(BT) is one of the widely used device for personal communication. As BT devices are operating in the unlicensed ISM band, they are often subjected to the interference from WLAN. The band width of BT (1MHz) is narrower compare to the bandwidth of WLAN (22MHz). So for coexistence purpose it is important to observe the performance of narrow band signal BT in presence of wideband interference WLAN and vice versa. As there are many work on the performance of WLAN in presence BT interference 3]4]5]6], the main focus in this paper is on performance of BT in presence of WLAN interference in AWGN, Rayleigh fading channel. Then comparison of the performance using interference avoidance technique like adaptive frequency hopping, power control for BT system is given. Finally a conclusion is drawn observing the simulation results on the technique which is more suitable for WLAN interference mitigation in BT system.