Speed control for low-energy on-time buses
<p>In 2015, 40% of the total energy used by final users in the United Kingdom was used in the transport sector. Hence it is important to minimise the energy consumption in the transport sector. This research focuses on the energy consumption minimisation of a bus using the approach of optim...
| Main Author: | |
|---|---|
| Other Authors: | |
| Format: | Thesis |
| Published: |
2017
|
| _version_ | 1797082540292440064 |
|---|---|
| author | Wong, R |
| author2 | McCulloch, M |
| author_facet | McCulloch, M Wong, R |
| author_sort | Wong, R |
| collection | OXFORD |
| description | <p>In 2015, 40% of the total energy used by final users in the United Kingdom was used in the transport sector. Hence it is important to minimise the energy consumption in the transport sector. This research focuses on the energy consumption minimisation of a bus using the approach of optimal control.</p> <p>The four objectives defined for this research are (i) the setting up of an optimal control problem for a bus journey, (ii) the formulation of a real-world bus route for open loop optimisation, (iii) the estimation of on-road bus speed for open loop optimisation, and (iv) the setting up of a real-time system to provide and update the optimal speed. An optimal control problem is set up to minimise the energy consumption and the discrepancy between the arrival time and the scheduled time at bus stops. The formulation of on-demand bus stops, compulsory bus stops, and slope changes is discussed. On-demand bus stops are formulated in the bost function, compulsory bus stops are formulated using multiphase approach, and the slope changes are interpolated using a smooth linear interpolation method. PSOPT, an open source optimal control program which implements the direct collocation method is used to solve the optimal control problem. The optimal profile of a basic route is compared with the profile resulted from an aggressive driving style. The optimal profile is found to consume 60% less energy than the aggressive driving profile.</p> <p>Two aspects are studied to incorporate a real-world bus route into the open loop optimisation process. The first aspect is the location-related aspect addressed by objective (ii). Methods to translate geodesic coordinates of a bus route and the bus stops are explored. The procedure to identify the stopping probabilities at bus stops is also discussed. The second aspect is the traffic-related aspect addressed by objective (iii). On-road traffic affects the ability of the bus to be optimally driven, therefore neural networks are designed to estimate the bus speed on both the mixed traffic lanes and the bus lanes.</p> <p>A real-time on-board system is set up to provide and update the optimal speed on the bus. The system acquires real-time information of the bus and adapts the optimal profile using a proposed adjustment algorithm when there are disruptions to the optimal driving. The adjustment algorithm found to be able to provide a near-optimal result with only 0.23% extra energy consumption compared to the optimal result with shorter execution time.</p> <p>This thesis uses a real-world bus route from Oxford City Centre to Pear Tree Park and Ride for case studies in different sections.</p> |
| first_indexed | 2024-03-07T01:29:23Z |
| format | Thesis |
| id | oxford-uuid:9317af03-3d37-41b6-9c2b-029130c3cde3 |
| institution | University of Oxford |
| last_indexed | 2024-03-07T01:29:23Z |
| publishDate | 2017 |
| record_format | dspace |
| spelling | oxford-uuid:9317af03-3d37-41b6-9c2b-029130c3cde32022-03-26T23:29:50ZSpeed control for low-energy on-time busesThesishttp://purl.org/coar/resource_type/c_db06uuid:9317af03-3d37-41b6-9c2b-029130c3cde3ORA Deposit2017Wong, RMcCulloch, M<p>In 2015, 40% of the total energy used by final users in the United Kingdom was used in the transport sector. Hence it is important to minimise the energy consumption in the transport sector. This research focuses on the energy consumption minimisation of a bus using the approach of optimal control.</p> <p>The four objectives defined for this research are (i) the setting up of an optimal control problem for a bus journey, (ii) the formulation of a real-world bus route for open loop optimisation, (iii) the estimation of on-road bus speed for open loop optimisation, and (iv) the setting up of a real-time system to provide and update the optimal speed. An optimal control problem is set up to minimise the energy consumption and the discrepancy between the arrival time and the scheduled time at bus stops. The formulation of on-demand bus stops, compulsory bus stops, and slope changes is discussed. On-demand bus stops are formulated in the bost function, compulsory bus stops are formulated using multiphase approach, and the slope changes are interpolated using a smooth linear interpolation method. PSOPT, an open source optimal control program which implements the direct collocation method is used to solve the optimal control problem. The optimal profile of a basic route is compared with the profile resulted from an aggressive driving style. The optimal profile is found to consume 60% less energy than the aggressive driving profile.</p> <p>Two aspects are studied to incorporate a real-world bus route into the open loop optimisation process. The first aspect is the location-related aspect addressed by objective (ii). Methods to translate geodesic coordinates of a bus route and the bus stops are explored. The procedure to identify the stopping probabilities at bus stops is also discussed. The second aspect is the traffic-related aspect addressed by objective (iii). On-road traffic affects the ability of the bus to be optimally driven, therefore neural networks are designed to estimate the bus speed on both the mixed traffic lanes and the bus lanes.</p> <p>A real-time on-board system is set up to provide and update the optimal speed on the bus. The system acquires real-time information of the bus and adapts the optimal profile using a proposed adjustment algorithm when there are disruptions to the optimal driving. The adjustment algorithm found to be able to provide a near-optimal result with only 0.23% extra energy consumption compared to the optimal result with shorter execution time.</p> <p>This thesis uses a real-world bus route from Oxford City Centre to Pear Tree Park and Ride for case studies in different sections.</p> |
| spellingShingle | Wong, R Speed control for low-energy on-time buses |
| title | Speed control for low-energy on-time buses |
| title_full | Speed control for low-energy on-time buses |
| title_fullStr | Speed control for low-energy on-time buses |
| title_full_unstemmed | Speed control for low-energy on-time buses |
| title_short | Speed control for low-energy on-time buses |
| title_sort | speed control for low energy on time buses |
| work_keys_str_mv | AT wongr speedcontrolforlowenergyontimebuses |