A novel adaptive weight selection algorithm for multi-objective multi-agent reinforcement learning

To solve multi-objective problems, multiple reward signals are often scalarized into a single value and further processed using established single-objective problem solving techniques. While the field of multi-objective optimization has made many advances in applying scalarization techniques to obtain good solution trade-offs, the utility of applying these techniques in the multi-objective multi-agent learning domain has not yet been thoroughly investigated. Agents learn the value of their decisions by linearly scalarizing their reward signals at the local level, while acceptable system wide behaviour results. However, the non-linear relationship between weighting parameters of the scalarization function and the learned policy makes the discovery of system wide trade-offs time consuming. Our first contribution is a thorough analysis of well known scalarization schemes within the multi-objective multi-agent reinforcement learning setup. The analysed approaches intelligently explore the weight-space in order to find a wider range of system trade-offs. In our second contribution, we propose a novel adaptive weight algorithm which interacts with the underlying local multi-objective solvers and allows for a better coverage of the Pareto front. Our third contribution is the experimental validation of our approach by learning bi-objective policies in self-organising smart camera networks. We note that our algorithm (i) explores the objective space faster on many problem instances, (ii) obtained solutions that exhibit a larger hypervolume, while (iii) acquiring a greater spread in the objective space.

Wireless machine-to-machine networks 2013

Editorial

Improving energy efficiency in a wireless sensor network by combining cooperative MIMO with data aggregation

In wireless sensor networks where nodes are powered by batteries, it is critical to prolong the network lifetime by minimizing the energy consumption of each node. In this paper, the cooperative multiple-input-multiple-output (MIMO) and data-aggregation techniques are jointly adopted to reduce the energy consumption per bit in wireless sensor networks by reducing the amount of data for transmission and better using network resources through cooperative communication. For this purpose, we derive a new energy model that considers the correlation between data generated by nodes and the distance between them for a cluster-based sensor network by employing the combined techniques. Using this model, the effect of the cluster size on the average energy consumption per node can be analyzed. It is shown that the energy efficiency of the network can significantly be enhanced in cooperative MIMO systems with data aggregation, compared with either cooperative MIMO systems without data aggregation or data-aggregation systems without cooperative MIMO, if sensor nodes are properly clusterized. Both centralized and distributed data-aggregation schemes for the cooperating nodes to exchange and compress their data are also proposed and appraised, which lead to diverse impacts of data correlation on the energy performance of the integrated cooperative MIMO and data-aggregation systems.

THE DESIGN OF A POWER SYSTEM FOR A WIRELESS SENSOR NETWORK FOR LONG-TERM OPERATION

Wireless sensor networks (WSNs) have shown wide applicability to many fields including monitoring of environmental, civil, and industrial settings. WSNs however are resource constrained by many competing factors that span their hardware, software, and networking. One of the central resource constrains is the charge consumption of WSN nodes. With finite energy supplies, low charge consumption is needed to ensure long lifetimes and success of WSNs. This thesis details the design of a power system to support long-term operation of WSNs. The power system’s development occurs in parallel with a custom WSN from the Queen’s MEMS Lab (QML-WSN), with the goal of supporting a 1+ year lifetime without sacrificing functionality. The final power system design utilizes a TPS62740 DC-DC converter with AA alkaline batteries to efficiently supply the nodes while providing battery monitoring functionality and an expansion slot for future development. Testing tools for measuring current draw and charge consumption were created along with analysis and processing software. Through their use charge consumption of the power system was drastically lowered and issues in QML-WSN were identified and resolved including the proper shutdown of accelerometers, and incorrect microcontroller unit (MCU) power pin connection. Controlled current profiling revealed unexpected behaviour of nodes and detailed current-voltage relationships. These relationships were utilized with a lifetime projection model to estimate a lifetime between 521-551 days, depending on the mode of operation. The power system and QML-WSN were tested over a long term trial lasting 272+ days in an industrial testbed to monitor an air compressor pump. Environmental factors were found to influence the behaviour of nodes leading to increased charge consumption, while a node in an office setting was still operating at the conclusion of the trail. This agrees with the lifetime projection and gives a strong indication that a 1+ year lifetime is achievable. Additionally, a light-weight charge consumption model was developed which allows charge consumption information of nodes in a distributed WSN to be monitored. This model was tested in a laboratory setting demonstrating +95% accuracy for high packet reception rate WSNs across varying data rates, battery supply capacities, and runtimes up to full battery depletion.

Techniques for batch reinforcement learning in robotics

This thesis addresses the Batch Reinforcement Learning methods in Robotics. This sub-class of Reinforcement Learning has shown promising results and has been the focus of recent research. Three contributions are proposed that aim to extend the state-of-art methods allowing for a faster and more stable learning process, such as required for learning in Robotics. The Q-learning update-rule is widely applied, since it allows to learn without the presence of a model of the environment. However, this update-rule is transition-based and does not take advantage of the underlying episodic structure of collected batch of interactions. The Q-Batch update-rule is proposed in this thesis, to process experiencies along the trajectories collected in the interaction phase. This allows a faster propagation of obtained rewards and penalties, resulting in faster and more robust learning. Non-parametric function approximations are explored, such as Gaussian Processes. This type of approximators allows to encode prior knowledge about the latent function, in the form of kernels, providing a higher level of exibility and accuracy. The application of Gaussian Processes in Batch Reinforcement Learning presented a higher performance in learning tasks than other function approximations used in the literature. Lastly, in order to extract more information from the experiences collected by the agent, model-learning techniques are incorporated to learn the system dynamics. In this way, it is possible to augment the set of collected experiences with experiences generated through planning using the learned models. Experiments were carried out mainly in simulation, with some tests carried out in a physical robotic platform. The obtained results show that the proposed approaches are able to outperform the classical Fitted Q Iteration.

Wireless Sensor Network Deployment

Wireless Sensor Networks (WSNs) are widely used for various civilian and military applications, and thus have attracted significant interest in recent years. This work investigates the important problem of optimal deployment of WSNs in terms of coverage and energy consumption. Five deployment algorithms are developed for maximal sensing range and minimal energy consumption in order to provide optimal sensing coverage and maximum lifetime. Also, all developed algorithms include self-healing capabilities in order to restore the operation of WSNs after a number of nodes have become inoperative. Two centralized optimization algorithms are developed, one based on Genetic Algorithms (GAs) and one based on Particle Swarm Optimization (PSO). Both optimization algorithms use powerful central nodes to calculate and obtain the global optimum outcomes. The GA is used to determine the optimal tradeoff between network coverage and overall distance travelled by fixed range sensors. The PSO algorithm is used to ensure 100% network coverage and minimize the energy consumed by mobile and range-adjustable sensors. Up to 30% - 90% energy savings can be provided in different scenarios by using the developed optimization algorithms thereby extending the lifetime of the sensor by 1.4 to 10 times. Three distributed optimization algorithms are also developed to relocate the sensors and optimize the coverage of networks with more stringent design and cost constraints. Each algorithm is cooperatively executed by all sensors to achieve better coverage. Two of our algorithms use the relative positions between sensors to optimize the coverage and energy savings. They provide 20% to 25% more energy savings than existing solutions. Our third algorithm is developed for networks without self-localization capabilities and supports the optimal deployment of such networks without requiring the use of expensive geolocation hardware or energy consuming localization algorithms. This is important for indoor monitoring applications since current localization algorithms cannot provide good accuracy for sensor relocation algorithms in such indoor environments. Also, no sensor redeployment algorithms, which can operate without self-localization systems, developed before our work.

Analysis and Evaluation of Tiny Machine Learning applications

The aim of TinyML is to bring the capability of Machine Learning to ultra-low-power devices, typically under a milliwatt, and with this it breaks the traditional power barrier that prevents the widely distributed machine intelligence. TinyML allows greater reactivity and privacy by conducting inference on the computer and near-sensor while avoiding the energy cost associated with wireless communication, which is far higher at this scale than that of computing. In addition, TinyML’s efficiency makes a class of smart, battery-powered, always-on applications that can revolutionize the collection and processing of data in real time. This emerging field, which is the end of a lot of innovation, is ready to speed up its growth in the coming years. In this thesis, we deploy three model on a microcontroller. For the model, datasets are retrieved from an online repository and are preprocessed as per our requirement. The model is then trained on the split of preprocessed data at its best to get the most accuracy out of it. Later the trained model is converted to C language to make it possible to deploy on the microcontroller. Finally, we take step towards incorporating the model into the microcontroller by implementing and evaluating an interface for the user to utilize the microcontroller’s sensors. In our thesis, we will have 4 chapters. The first will give us an introduction of TinyML. The second chapter will help setup the TinyML Environment. The third chapter will be about a major use of TinyML in Wake Word Detection. The final chapter will deal with Gesture Recognition in TinyML.

Thermal and air quality modeling of an electric vehicle cabin with low-cost sensors and reinforcement learning

The design process of any electric vehicle system has to be oriented towards the best energy efficiency, together with the constraint of maintaining comfort in the vehicle cabin. Main aim of this study is to research the best thermal management solution in terms of HVAC efficiency without compromising occupant’s comfort and internal air quality. An Arduino controlled Low Cost System of Sensors was developed and compared against reference instrumentation (average R-squared of 0.92) and then used to characterise the vehicle cabin in real parking and driving conditions trials. Data on the energy use of the HVAC was retrieved from the car On-Board Diagnostic port. Energy savings using recirculation can reach 30 %, but pollutants concentration in the cabin builds up in this operating mode. Moreover, the temperature profile appeared strongly nonuniform with air temperature differences up to 10° C. Optimisation methods often require a high number of runs to find the optimal configuration of the system. Fast models proved to be beneficial for these task, while CFD-1D model are usually slower despite the higher level of detail provided. In this work, the collected dataset was used to train a fast ML model of both cabin and HVAC using linear regression. Average scaled RMSE over all trials is 0.4 %, while computation time is 0.0077 ms for each second of simulated time on a laptop computer. Finally, a reinforcement learning environment was built in OpenAI and Stable-Baselines3 using the built-in Proximal Policy Optimisation algorithm to update the policy and seek for the best compromise between comfort, air quality and energy reward terms. The learning curves show an oscillating behaviour overall, with only 2 experiments behaving as expected even if too slow. This result leaves large room for improvement, ranging from the reward function engineering to the expansion of the ML model.

Dynamic equilibrium through reinforcement learning

Reinforcement Learning is an area of Machine Learning that deals with how an agent should take actions in an environment such as to maximize the notion of accumulated reward. This type of learning is inspired by the way humans learn and has led to the creation of various algorithms for reinforcement learning. These algorithms focus on the way in which an agent’s behaviour can be improved, assuming independence as to their surroundings. The current work studies the application of reinforcement learning methods to solve the inverted pendulum problem. The importance of the variability of the environment (factors that are external to the agent) on the execution of reinforcement learning agents is studied by using a model that seeks to obtain equilibrium (stability) through dynamism – a Cart-Pole system or inverted pendulum. We sought to improve the behaviour of the autonomous agents by changing the information passed to them, while maintaining the agent’s internal parameters constant (learning rate, discount factors, decay rate, etc.), instead of the classical approach of tuning the agent’s internal parameters. The influence of changes on the state set and the action set on an agent’s capability to solve the Cart-pole problem was studied. We have studied typical behaviour of reinforcement learning agents applied to the classic BOXES model and a new form of characterizing the environment was proposed using the notion of convergence towards a reference value. We demonstrate the gain in performance of this new method applied to a Q-Learning agent.

Intelligent decision making in electricity markets: simulated annealing Q-Learning

Electricity markets are complex environments, involving a large number of different entities, playing in a dynamic scene to obtain the best advantages and profits. MASCEM is a multi-agent electricity market simulator to model market players and simulate their operation in the market. Market players are entities with specific characteristics and objectives, making their decisions and interacting with other players. MASCEM is integrated with ALBidS, a system that provides several dynamic strategies for agents’ behavior. This paper presents a method that aims at enhancing ALBidS competence in endowing market players with adequate strategic bidding capabilities, allowing them to obtain the higher possible gains out of the market. This method uses a reinforcement learning algorithm to learn from experience how to choose the best from a set of possible actions. These actions are defined accordingly to the most probable points of bidding success. With the purpose of accelerating the convergence process, a simulated annealing based algorithm is included.

Reinforcement Learning Based on the Bayesian Theorem for Electricity Markets Decision Support

This paper presents the applicability of a reinforcement learning algorithm based on the application of the Bayesian theorem of probability. The proposed reinforcement learning algorithm is an advantageous and indispensable tool for ALBidS (Adaptive Learning strategic Bidding System), a multi-agent system that has the purpose of providing decision support to electricity market negotiating players. ALBidS uses a set of different strategies for providing decision support to market players. These strategies are used accordingly to their probability of success for each different context. The approach proposed in this paper uses a Bayesian network for deciding the most probably successful action at each time, depending on past events. The performance of the proposed methodology is tested using electricity market simulations in MASCEM (Multi-Agent Simulator of Competitive Electricity Markets). MASCEM provides the means for simulating a real electricity market environment, based on real data from real electricity market operators.

The Visual and Virtual in Contemporary European Learning Contexts

Special issue of Anthropology in Action originated from the Working Images Conference, a joint meeting of TAN and VAN EASA networks

Performance assessment of electrically small antennas for implantable microsystems with wireless power and communications

Dissertação de mestrado integrado em Engenharia Biomédica (área de especialização em Eletrónica Médica)