Page - (000421) - in Autonomes Fahren - Technische, rechtliche und gesellschaftliche Aspekte

Autonomous Mobility-on-Demand Systems for Future Urban Mobility406 can be cast as a linear program (hence, this approach extends well to large transportation networks). This method encourages coordination but does not enforce it, which is the key to maintaining tractability of the model [13]. The rates {ȥi} and probabilities {Įi j} are then used as feedforward reference signals in a receding horizon control scheme to control in real-time an entire AMoD system [13], as done for case studies of New York City and Singapore presented in Section 19.3. 19.2.2.2 Distributed approach The key idea behind the distributed approach [14, 15, 16, 17] is that the number of stations represents a continuum (i.e., N ĺ ’ VLPLODU WR WKH '\QDPLF 7UDYHOLQJ 5HSDLUPDQ SUREOHP [20, 21, 22, 23]. In other words, customers arrive at any point in a given bounded environ- ment [15, 16], or at any point along the segments of a road map [15]. In the simplest sce- nario, a dynamical process generates spatially-localized origin-destination pairs (represent- ing the transportation requests) in a geographical region Q  R2 . The process that generates origin-destination pairs is modeled as a spatio-temporal Poisson process, namely, (i) the time between consecutive generation instants has an exponential distribution with intensity Ȝ , and (ii) origins and destinations are random variables with probability density functions, respectively, ijO and ijD , supported over Q, see Figure 19.3 (right). The objective is to design a routing policy that minimizes the average steady-state time delay between the generation of an origin-destination pair and the time the trip is completed. By removing WKH FRQVWUDLQW WKDW FXVWRPHUV¶ RULJLQ GHVWLQDWLRQ SDLUV DUH ORFDOL]HG DW D ¿QLWH VHW RI SRLQWV in an environment, one transforms the problem of controlling N different queues into one RI FRQWUROOLQJ D VLQJOH ³VSDWLDOO\ DYHUDJHG´ TXHXH 7KLV FRQVLGHUDEO\ VLPSOL¿HV DQDO\VLV and control, and allows one to derive analytical expressions for important design parame- Fig. 19.3 Left figure: In the lumped model, an AMoD system is modeled as a Jackson network, where stations are identified with single-server queues and roads are identified with infinite-server queues. (Customers are denoted with yellow dots and servicing vehicles are represented by small car icons.) Some vehicles travel without passengers to rebalance the fleet. Right figure: In a distributed model of an AMoD system, a stochastic process with rate Ȝ generates origin-destination pairs, dis- tributed over a continuous domain Q.