Carlo Manna

In-Time Prognosis Based on Swarm Intelligence for Home-Care Monitoring: A Case Study on Pulmonary Disease

A swarm intelligence-based procedure to detect critical conditions of a patient, affected by a specific disease, at an early stage in absence of clinician, is proposed. The procedure is to be integrated inside a remote health care system for patients at home, where some physiological parameters related to a specific disease are being monitored. A significant variation in the monitored parameters can lead the patient to a critical state, thus the proposed method is aimed at predicting a possible future bad condition of the patient on the basis of past measurements. Moreover, different physiological parameters contribute to diverse degrees in dissimilar diseases; consequently, a swarm intelligence-based method is proposed for optimizing the weight of each parameter for a more accurate diagnosis. The proposed approach has been validated experimentally under the framework of the industrial research project Patient Diagnosis and Monitoring at Domicile (PADIAMOND: co-funded by EU and the company Filia srl, Caserta, Italy).

A Quantum-inspired Evolutionary Algorithm with a competitive variation operator for Multiple-Fault Diagnosis

A heuristic search algorithm, the Quantum-inspired Competitive Evolutionary Algorithm (QuCEA), based on both quantum and evolutionary computing, is proposed. The individuals of a population, coded as qubit strings, evolve by means of an original variation operator inspired by competitive learning. The proposed operator is application independent and intuitively controllable by a single real parameter. QuCEA has been applied to Multiple-Fault Diagnosis, a typical NP-hard problem for industrial diagnosis. In particular, the proposed algorithm gives remarkable results both in simulation and in on-field tests for a lift monitoring system, also in comparison with a standard genetic algorithm and a state-of-the-art Quantum-inspired Evolutionary Algorithm.

Online Stochastic Planning for Taxi and Ridesharing

In this paper we consider the problem of on-line stochastic ride-sharing and taxi-sharing with time windows. We study a scenario in which people needing a taxi, or a ride, assign their source and destination points plus other restrictions (such as earlier time to departure and maximum time to reach a destination), at the same time, there are taxis or drivers interested in providing a ride (also with departure and destination points, vehicle capacity and time restrictions). We model the time window restrictions as a soft constraint (a reasonable delay might be acceptable in a realistic scenario), and consider the problem as an on-line continual planning problem, in which additional ride requests may arrive while plans for previous ride-matching are being executed. Finally, such new requests may arrive at each time step with some probability. The aim is to maximize the shared trips while minimising the expected travel delay for each trip. In this paper we propose an on-line stochastic optimization planning approach in which instead of myopically optimising for the offered trips and requested trips that are known, incorporate information that partially describes the stochastic future into the model in order to improve the quality of the solution. We prove the effectiveness of the method in a real world scenario using a number of instances extracted from a travel survey in north-eastern Illinois (USA) conducted by the Chicago Metropolitan Agency for Planning.

Learning Occupancy in Single Person Offices with Mixtures of Multi-lag Markov Chains

The problem of real-time occupancy forecasting for single person offices is critical for energy efficient buildings which use predictive control techniques. Due to the highly uncertain nature of occupancy dynamics, the modeling and prediction of occupancy is a challenging problem. This paper proposes an algorithm for learning and predicting single occupant presence in office buildings, by considering the occupant behaviour as an ensemble of multiple Markov models at different time lags. This model has been tested using real occupancy data collected from PIR sensors installed in three different buildings and compared with state of the art methods, reducing the error rate by on average 5% over the best comparator method.

Ant-Search Strategy Based on Likelihood Trail Intensity Modification for Multiple-Fault Diagnosis in Sensor Networks

A swarm intelligence-based approach to multiple-fault diagnostics for industrial applications is proposed. Drawbacks of swarm-based algorithms in heuristic search strategy related to mutual dependence of solutions are overcome by a likelihood-based trail intensity modification of ant-colony optimization. Numerical results of a comprehensive characterization through statistical experiment design on high-dimension multiple-faults diagnosis applications are shown. Experimental results under the framework of an industrial research project committed to industrial remote monitoring of operating machines are discussed. Numerical and experimental results show excellent performance, outperforming genetic algorithms, especially in high-dimension problems, and easiness in algorithm configuration.

Model predictive control strategy based on differential discrete particle swarm optimization

A model-based predictive control strategy, aided by a differential discrete particle swarm optimization, is proposed. In particular, the proposed approach extends the traditional discrete binary version of the particle swarm algorithm by redesigning the particle section (all possible solutions in the search space) in order to represent a sequence of discrete controls. In this way, the “velocity section” turns out to be related to the probability of achieving the suitable discrete control value. As an experimental case study, this strategy is applied to a temperature control in a building aimed at energy saving. The proposed method is compared with a standard particle swarm algorithm, and experimental results are discussed.

On-line Dynamic Station Redeployments in Bike-Sharing Systems

Bike-sharing has seen great development during recent years, both in Europe and globally. However, these systems are far from perfect. The uncertainty of the customer demand often leads to an unbalanced distribution of bicycles over the time and space congestion and/or starvation, resulting both in a loss of customers and a poor customer experience. In order to improve those aspects, we propose a dynamic bike-sharing system, which combines the standard fixed base stations with movable stations using trucks, which will able to be dynamically re-allocated according to the upcoming forecasted customer demand during the day in real-time. The purpose of this paper is to investigate whether using moveable stations in designing the bike-sharing system has a significant positive effect on the system performance. To that end, we contribute an on-line stochastic optimization formulation to address the redeployment of the moveable stations during the day, to better match the upcoming customer demand. Finally, we demonstrate the utility of our approach with numerical experiments using data provided by bike-sharing companies.

Personalized Thermal Comfort Forecasting for Smart Buildings via Locally Weighted Regression with Adaptive Bandwidth

Irish Research Council for Science, Engineering and Technology (Enterprise Partnership Scheme)

Automatic modeling based on cultural programming for osseointegration diagnosis

The problem of modeling equivalent circuits for interpreting Electrical Impedance Spectroscopy (EIS) data in monitoring osseointegration level of metallic implants in bone is faced by means of an evolutionary programming approach based on cultural algorithms. With respect to state-of-the-art gene expression programming, the information on search advance acquired by most promising individuals during the evolution is shared with the entire population of potential solutions and stored also for next generations. Experimental results of the application such cultural programming-based analytical modeling to in-vitro EIS measurements of bone in-growth around metallic implants during prosthesis osseointegration are presented.