Safety is one of the most important requirements in the transportation world. The reduction of the number of road accidents is an always up-to-date argument. In the last decades, the number of fatalities was almost halved together with the related social costs, but still today the accidents number is impressive (1 million only in Europe in 2015). Increasing the road safety is an ethical challenge that could save human lives and reduce social expenditures. In this thesis, two kinds of force sensors for detecting loads acting on vehicles have been developed and tested. The first family of sensors refers to intelligent wheels, able to measure three forces and three moments acting at the hub. The wheels are motorcycle type, particularly suited for future urban light vehicles. The second kind of sensor is a steering wheel, particularly suited for Advanced Driver Assistance Systems studies. Starting from a Politecnico di Milano patent, the sensors have been developed up to actual production and professional employment. The development of the sensors focused primarily (but not only) on an optimization algorithm that was developed to define the best geometry for optimal measuring performance. All of the sensor components have been designed with a 3D CAD modeller and verified by means of a Finite Element Analysis. The sensors have been instrumented with twelve strain gauges each and experimentally calibrated to obtain the best measuring performance. Finally, the force sensors have been experimentally tested and used on real vehicles, either motorcycles or cars. Two different smart wheels for motorcycle have been developed, that allow the estimation of the contact forces between the motorcycle and the road. After the construction and the experimental calibration, the smart wheels have been used on a race motorcycle during a track test. The acquired data were used to derive combined load spectra with an innovative method. The track is split in several well defined manoeuvre patterns, like the straight acceleration or braking, the acceleration or braking in exiting or entering a curve, the steady state turning, the curb hitting and the gear-shifting. New combined load spectra are evaluated for the motorcycle, useful in the design phase of the new components. The second force sensor developed is an innovative Instrumented Steering Wheel that measures the forces exerted by each hand of the driver. A carbon fibre reinforced composite structure is fitted with two original six-axis load cells. The carbon fibre reinforced structure has been modelled with a Finite Element method to verify the stress state and the eigen-frequencies. The load cells have been instrumented with twelve strain gauges each and experimentally calibrated. Finally, the Instrumented Steering Wheel has been tested on a track. The vehicle passed by a kick-plate that induced over-steer and eventually spin if the driver did not react properly. The data from nine different drivers were acquired. In order to obtain the highest possible sensitivity, the inertial force contributions acting on the steering wheel structure have been compensated. Analysing the acquired data, some common driving behaviours among the drivers have been highlighted, either for a classic turning manoeuvre, or for an emergency manoeuvre. Moreover, forces variations have been noticed before the actual rotation of the steering wheel in the emergency manoeuvre, before the yaw acceleration. This occurrence could be useful in tuning properly the ADAS to shorten the intervention time and increase safety. A new approach is presented to find new KPIs for the steering evaluation, analysing the different forces applied by the driver with the introduction of the ISW. Finding a complete correlation between the driver feeling and current objective indexes is typically difficult to achieve. Existing analysis techniques generate KPIs which are based on the vehicle and the steering dynamics magnitudes: lateral acceleration, steering torque, steering angle, etc. However, driver action is only partially captured by the resulting steering torque. New KPIs are related to the steering efficiency, defined as the ratio between the tangential steering forces and the total forces. Moreover, the knowledge of the loads exerted by the driver could lead to a deeper knowledge of the steering behaviour: the loads applied by the driver could change with the vehicle response. The driver should be considered in the steering loop. Both of the two force sensors allow to improve the knowledge of vehicle interaction with the road and with the driver in order to increase safety in Intelligent Transport System framework.
Sicurezza è uno dei concetti più importanti nel mondo dei trasporti: la riduzione del numero di incidenti stradali è un argomento sempre in via di evoluzione. Negli ultimi decenni il numero di morti dovuti ad incidenti stradali è stato dimezzato, insieme con i relativi costi sociali, ma tutt'oggi il numero di incidenti stradali è impressionante (un milione in Europa solo nel 2015). Aumentare la sicurezza stradale è una sfida etica che può salvare vite umane e ridurre le spese sociali. In questa tesi sono stati sviluppati e testati due diversi tipi di sensori per la misura delle forze agenti sui veicoli. Il primo tipo di sensori sono delle ruote strumentate, capaci di misurare le tre componenti di forza e di momento agenti al mozzo della ruota stessa. Le ruote sviluppate sono per applicazioni motociclistiche, particolarmente indicate per i futuri veicoli leggeri per utilizzo urbano. Il secondo tipo di sensore è un volante strumentato, particolarmente indicato per lo sviluppo dei sistemi di aiuto al conducente (ADAS). I sensori sono stati sviluppati partendo da un brevetto del Politecnico di Milano, fino ad arrivare alla loro realizzazione e al loro utilizzo professionale. È stato sviluppato un algoritmo di ottimizzazione per guidare i progettisti nella fase di design dei sensori, per individuare la migliore geometria che porti alle migliori performance di misura. Tutti i sensori sono stati modellati con un CAD 3D e studiati tramite analisi agli elementi finiti. Ogni sensore è stato strumentato con dodici estensimetri e calibrato sperimentalmente per ottenere le migliori performance di misura. Infine, i sensori sono stati testati sperimentalmente e adottati sui veicoli, sia sulle motociclette che sulle automobili. Sono state progettate due diverse ruote strumentate, che permettono la stima delle forze di contatto tra la motocicletta e la strada. Dopo la loro realizzazione e la loro calibrazione sperimentale, le due ruote strumentate sono state usate su una motocicletta da competizione durante un test su pista. I dati ottenuti sono stati utilizzati per il calcolo di spettri di carico combinati per uso motociclistico con una nuova metodologia proposta. Diversi pattern di forza sono stati individuati, uno per ogni manovra come l'accelerazione o frenata in rettilineo, l'accelerazione in uscita curva, la frenata in ingresso curva, la curva piena, il cordolo e il cambio marcia. Sono stati ottenuti nuovi spettri di carico combinati per uso motociclistico, utilizzabili nella fase di progettazione di nuovi componenti motociclistici. È stato progettato un volante strumentato, che permette la misura delle forze e dei momenti applicati da ogni mano del conducente. Il corpo del volante è in composito di fibra di carbonio, ed è strumentato con due celle di carico a sei componenti. La struttura in composito è stata modellata con gli elementi finiti per valutarne lo stato di sforzo e le frequenze proprie. Infine, le celle di carico sono state strumentate con dodici estensimetri ognuna e sono state calibrate sperimentalmente. Il volante strumentato è stato usato durante un test in pista. L'automobile doveva passare sopra l'impianto del kick plate che, spostandosi lateralmente, simula un disturbo laterale sull'asse posteriore del veicolo, che porta ad un controsterzo ed eventualmente ad un testacoda se il conducente non reagisce in modo opportuno. Sono stati analizzati i dati di nove diversi conducenti. Per ottenere la massima sensitività possibile, i contributi di forza inerziale che agiscono sul volante sono stati opportunamente compensati. Analizzando i dati acquisiti, sono state individuate alcune strategie di guida comuni tra i vari conducenti, sia analizzando le manovre a bassa velocità che analizzando le manovre di emergenza. Inoltre, una variazione di forza sul volante viene applicata da tutti i conducenti prima della effettiva rotazione del volante durante la manovra di emergenza. Questa applicazione di carico sul volante può essere sfruttata per tarate opportunamente gli ADAS per accorciarne il tempo di intervento e aumentare la sicurezza stradale. Viene, inoltre, presentato un nuovo approccio per trovare nuovi KPIs per la valutazione dello sterzo, analizzando le diverse forze fatte dal conducente sul volante strumentato. Trovare una correlazione tra le sensazioni soggettive del conducente e degli indici oggettivi è estremamente difficile. In letteratura ci sono alcuni indici basati sulle variabili dinamiche del veicolo, come l'accelerazione laterale, la coppia di sterzo e l'angolo di sterzo. Purtroppo, l'azione del conducente si riflette solo parzialmente sulla coppia di sterzo applicata. In questa tesi vengono proposti nuovi KPIs relativi alla steering efficiency, definita come rapporto tra le forze tangenziali e le altre componenti di forza applicate dal conducente. Infatti, conoscere le forze applicate dal conducente sul volante strumentato potrebbe portare ad una migliore conoscenza dell'interazione tra l'uomo e il veicolo: le forze applicate dal conducente cambiano la risposta del veicolo, vanno quindi considerate nell'analisi del fenomeno. Entrambi i sensori permettono di incrementare la conoscenza dell'interazione del veicolo con la strada e con il conducente, per aumentare la sicurezza stradale nell'ambito del sistema dei trasporti intelligenti (ITS).
Force sensors as smart vehicle subsystems to improve safety in Intelligent Transport Systems (ITS)
FRANCESCO, COMOLLI
2019
Abstract
Safety is one of the most important requirements in the transportation world. The reduction of the number of road accidents is an always up-to-date argument. In the last decades, the number of fatalities was almost halved together with the related social costs, but still today the accidents number is impressive (1 million only in Europe in 2015). Increasing the road safety is an ethical challenge that could save human lives and reduce social expenditures. In this thesis, two kinds of force sensors for detecting loads acting on vehicles have been developed and tested. The first family of sensors refers to intelligent wheels, able to measure three forces and three moments acting at the hub. The wheels are motorcycle type, particularly suited for future urban light vehicles. The second kind of sensor is a steering wheel, particularly suited for Advanced Driver Assistance Systems studies. Starting from a Politecnico di Milano patent, the sensors have been developed up to actual production and professional employment. The development of the sensors focused primarily (but not only) on an optimization algorithm that was developed to define the best geometry for optimal measuring performance. All of the sensor components have been designed with a 3D CAD modeller and verified by means of a Finite Element Analysis. The sensors have been instrumented with twelve strain gauges each and experimentally calibrated to obtain the best measuring performance. Finally, the force sensors have been experimentally tested and used on real vehicles, either motorcycles or cars. Two different smart wheels for motorcycle have been developed, that allow the estimation of the contact forces between the motorcycle and the road. After the construction and the experimental calibration, the smart wheels have been used on a race motorcycle during a track test. The acquired data were used to derive combined load spectra with an innovative method. The track is split in several well defined manoeuvre patterns, like the straight acceleration or braking, the acceleration or braking in exiting or entering a curve, the steady state turning, the curb hitting and the gear-shifting. New combined load spectra are evaluated for the motorcycle, useful in the design phase of the new components. The second force sensor developed is an innovative Instrumented Steering Wheel that measures the forces exerted by each hand of the driver. A carbon fibre reinforced composite structure is fitted with two original six-axis load cells. The carbon fibre reinforced structure has been modelled with a Finite Element method to verify the stress state and the eigen-frequencies. The load cells have been instrumented with twelve strain gauges each and experimentally calibrated. Finally, the Instrumented Steering Wheel has been tested on a track. The vehicle passed by a kick-plate that induced over-steer and eventually spin if the driver did not react properly. The data from nine different drivers were acquired. In order to obtain the highest possible sensitivity, the inertial force contributions acting on the steering wheel structure have been compensated. Analysing the acquired data, some common driving behaviours among the drivers have been highlighted, either for a classic turning manoeuvre, or for an emergency manoeuvre. Moreover, forces variations have been noticed before the actual rotation of the steering wheel in the emergency manoeuvre, before the yaw acceleration. This occurrence could be useful in tuning properly the ADAS to shorten the intervention time and increase safety. A new approach is presented to find new KPIs for the steering evaluation, analysing the different forces applied by the driver with the introduction of the ISW. Finding a complete correlation between the driver feeling and current objective indexes is typically difficult to achieve. Existing analysis techniques generate KPIs which are based on the vehicle and the steering dynamics magnitudes: lateral acceleration, steering torque, steering angle, etc. However, driver action is only partially captured by the resulting steering torque. New KPIs are related to the steering efficiency, defined as the ratio between the tangential steering forces and the total forces. Moreover, the knowledge of the loads exerted by the driver could lead to a deeper knowledge of the steering behaviour: the loads applied by the driver could change with the vehicle response. The driver should be considered in the steering loop. Both of the two force sensors allow to improve the knowledge of vehicle interaction with the road and with the driver in order to increase safety in Intelligent Transport System framework.File | Dimensione | Formato | |
---|---|---|---|
PhD_thesis.pdf
accesso solo da BNCF e BNCR
Dimensione
19.89 MB
Formato
Adobe PDF
|
19.89 MB | Adobe PDF |
I documenti in UNITESI sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
https://hdl.handle.net/20.500.14242/205597
URN:NBN:IT:POLIMI-205597