Luis Alejandro Sánchez-Pérez

Aircraft class identification based on take-off noise signal segmentation in time

Abstract Aircraft noise is one of the most uncomfortable kinds of sounds. That is why many organizations have addressed this problem through noise contours around airports, for which they use the aircraft type as the key element. This paper presents a new computational model to identify the aircraft class with a better performance, because it introduces the take-off noise signal segmentation in time. A method for signal segmentation into four segments was created. The aircraft noise patterns are extracted using an LPC (Linear Predictive Coding) based technique and the classification is made combining the output of four parallel MLP (Multilayer Perceptron) neural networks, one for each segment. The individual accuracy of each network was improved using a wrapper feature selection method, increasing the model effectiveness with a lower computational cost. The aircraft are grouped into classes depending on the installed engine type. The model works with 13 aircraft categories with an identification level above 85% in real environments.

Geo-referenced flight path estimation based on spatio-temporal information extracted from aircraft take-off noise

Abstract The closer proximity between airports and residential areas has created a growing attention regarding noise pollution. The noise abatement procedures established by the aeronautical authorities and the models for computing noise contours around airports are proof of that. There are also models for identifying aircraft taking off which have focused on the correlation between the aircraft position and the noise signal. However, this correlation has been made so far without spatial information. The present study proposes a method to estimate the geo-referenced flight path followed by an aircraft taking off, using the spatio-temporal information extracted from the noise signal and improved with a smoothing algorithm. A microphone array with twelve sensors is used in order to evaluate different sensor spacings and the spatial aliasing effect when working with take-off noise signals. The flight path estimation method assumes that the aircraft is following a ground track collinear to the runway and was compared against radar information and Automatic Dependent Surveillance-Broadcast (ADS-B) data. The average method accuracy was between 3 and 6 meters. The estimated flight path has a ground length of about two kilometers, including locations at least one kilometer apart from the measurement point.

Classification of unbalance and misalignment in induction motors using orbital analysis and associative memories

Fault detection in induction motors is an important task in industry when production greatly depends of the functioning of the machine. This paper presents a new computational model for detecting misalignment and unbalance problems in electrical induction motors. Through orbital analysis and signal vibrations, unbalance and misalignment motor faults can be mapped into patterns, which are processed by a classifier: the Steinbuch Lernmatrix. This associative memory has been widely used as classifier in the pattern recognition field. A modification of the Lernmatrix is proposed in order to process real valued data and improve the efficiency and performance of the classifier. Experimental patterns obtained from induction motors in real situations and with a certain level of unbalance or misalignment were processed by the proposed model. Classification results obtained in an experimental phase indicate a good performance of the associative memory, providing an alternative way for recognizing induction motor faults.

Dynamic hierarchical aggregation of parallel outputs for aircraft take-off noise identification

Assessment of airport noise pollution mainly depends on the correlation between aircraft class, noise measured and flight path geometry. Regulation, evaluation and especially certification procedures generally establish that previous correlation cannot be carried out using aircraft navigation systems data. Additionally, airport noise monitoring systems generally use aircraft noise signals only for computing statistical indicators. Consequently, methods to acquire more information from these signals have been explored so as to improve noise estimation around airports. In this regard, this paper introduces a new model for aircraft class recognition based on take-off noise signal segmentation and dynamic hierarchical aggregation of K parallel neural networks outputs O p k . A single hierarchy is separately defined for every class p, mainly based on the recall and precision of neural network NNk|k=1,2,?,K. Similarly, the dynamics proposed is also particular to each class p. The performance of the new model is benchmarked against models in literature over a database containing real-world take-off noise measurements. The new model performs better on the abovementioned database and successfully classifies over 89% of measurements.

Airport take-off noise assessment aimed at identify responsible aircraft classes.

Assessment of aircraft noise is an important task of nowadays airports in order to fight environmental noise pollution given the recent discoveries on the exposure negative effects on human health. Noise monitoring and estimation around airports mostly use aircraft noise signals only for computing statistical indicators and depends on additional data sources so as to determine required inputs such as the aircraft class responsible for noise pollution. In this sense, the noise monitoring and estimation systems have been tried to improve by creating methods for obtaining more information from aircraft noise signals, especially real-time aircraft class recognition. Consequently, this paper proposes a multilayer neural-fuzzy model for aircraft class recognition based on take-off noise signal segmentation. It uses a fuzzy inference system to build a final response for each class p based on the aggregation of K parallel neural networks outputs Op(k) with respect to Linear Predictive Coding (LPC) features extracted from K adjacent signal segments. Based on extensive experiments over two databases with real-time take-off noise measurements, the proposed model performs better than other methods in literature, particularly when aircraft classes are strongly correlated to each other. A new strictly cross-checked database is introduced including more complex classes and real-time take-off noise measurements from modern aircrafts. The new model is at least 5% more accurate with respect to previous database and successfully classifies 87% of measurements in the new database.

Computer model for leg agility quantification and assessment for Parkinson’s disease patients

AbstractParkinson’s disease (PD) is a progressive disorder that affects motor regulation. The Unified Parkinson’s Disease Rating Scale sponsored by the Movement Disorder Society (MDS-UPDRS) quantifies the illness progression based on clinical observations. The leg agility is an item in this scale, yet only a visual detection of the features is used, leading to subjectivity. Overall, 50 patients (85 measurements) with varying motor impairment severity were asked to perform the leg agility item while wearing inertial sensor units on each ankle. We quantified features based on the MDS-UPDRS and designed a fuzzy inference model to capture clinical knowledge for assessment. The model proposed is capable of capturing all details regardless of the task speed, reducing the inherent uncertainty of the examiner observations obtaining a 92.35% of coincidence with at least one expert. In addition, the continuous scale implemented in this work prevents the inherent “floor/ceil” effect of discrete scales. This model proves the feasibility of quantification and assessment of the leg agility through inertial signals. Moreover, it allows a better follow-up of the PD patient state, due to the repeatability of our computer model and the continuous output, which are not objectively achievable through visual examination. Graphical abstractᅟ

A Modification of the Lernmatrix for Real Valued Data Processing

An associative memory is a binary relationship between inputs and outputs, which is stored in an M matrix. In this paper, we propose a modification of the Steinbuch Lernmatrix model in order to process real-valued patterns, avoiding binarization processes and reducing computational burden. The proposed model is used in experiments with noisy environments, where the performance and efficiency of the memory is proven. A comparison between the proposed and the original model shows a good response and efficiency in the classification process of the new Lernmatrix.