JP4417329B2 - Elevator delivery that adjusts passenger sensory waiting time - Google Patents

Description

  The present invention provides elevator delivery that adjusts the passenger's sensory waiting time until the elevator arrives to the travel time in which the passenger in the elevator feels shorter, and the lowest total sensory service time. Concerning the choice of assignment to a call that gives a function.

  It has been found that in the delivery of an elevator, the passenger's sense of how long it waited for the elevator to arrive is non-linear. That is, the passenger feels that the longer the waiting time, the longer the waiting time than the actual waiting time. In other words, the degree of distress due to waiting is not a linear function of waiting time, but probably increases exponentially over time. In US Pat. No. 5,304,752, service is preferentially allocated to the longest waiting time among all passengers currently waiting for the elevator to arrive. In U.S. Pat. No. 4,244,450, in order to deliver a car more dependent on the passenger's sensory waiting time, the delivery device uses an increasing function of waiting time that increases over time. In this patent the allocation is based on the minimum of the sum of all sensory waiting times of all waiting passengers. Many systems are equipped with a display device that displays the remaining time for a call, giving passengers a sense of security that they will be available soon. One example is US Pat. No. 5,789,715.

  It is an object of the present invention to improve passenger waiting and service time perception in an elevator system, deliver elevators in response to hall calls in a manner that makes passengers feel the pain of waiting to a minimum, and The method of responding to and service of elevator calls includes delivering the elevator so as to obtain more passenger support.

  The first aspect of the present invention is that the elevator passengers who rush forward feel that the waiting time is longer than the sensual service time, that is, the time spent waiting for the arrival of the elevator car, rather than the traveling time to the destination floor. Based on our discovery that it is more serious. The present invention also provides that the total passive time of all passengers during the elevator service is controlled by the least square sum of other sensory delays or other exponential functions until all passengers reach the destination floor. If so, it is based in part on our discovery that it feels shorter.

  According to the present invention, a sensory waiting time that is a function of an expected time until a specific car can handle a specific call, a travel time to the destination floor (estimated time or real time), and a fixed time are determined. And the sensory driving time, which is a function of the weighted value of the time corresponding to the scheduled stop, gives a sensory service time, where the waiting time before service is sensory Is weighted more than the average travel time. Furthermore, according to the invention, for all possible sets of allocations for all unsupported up and down hall calls currently waiting in the elevator system, a sensory service for all unsupported hall calls. The time (waiting time and travel time) is squared and the squares are summed. The set of assignments that gives the minimum sum of squares is the set on which the car assignment corresponding to the call should be based.

  The present invention can use a neural network to determine some components of the factors on which the allocation for calls should be based. Specifically, the remaining corresponding time, commonly referred to as RRT, may be determined by a neural network as disclosed in US Pat. No. 5,672,853. Other predictions, such as predicted travel time for each car, and the anticipated new stop are also used, if desired, using the neural network processing method disclosed in US Pat. No. 5,672,853. It may be determined.

  Other objects, features and benefits of the present invention will become apparent from the detailed description of the examples below.

  Referring to FIG. 1, the signal processing device 11 shows a group control device that allocates a car to correspond to a hall call using the aspect of the present invention. The processing device 11 responds to a plurality of sensors 12 such as a car weight sensor provided for the input / output (I / O) port 15 of the processing apparatus 11 and a data signal 13 such as car direction and door status. Similarly, other I / O ports 18 have a plurality of hall call buttons 19 placed on various floors of the building, a plurality of in-car call button panels 20 placed on each car, and usually at a boarding place on each floor. One or two or more hall indicator lights 21 are connected. The processing unit 11 includes a data bus 24, an address bus 25, a central processing unit (CPU) 26, a random access memory (RAM) 27, and a read-only memory that stores essential elements of a program or routine that can implement the present invention. ROM) 28.

  In the routine of FIG. 2, as will be described, an upstream hall call is processed first, then a downstream hall call is processed, and then all calls are processed together.

  Referring to FIG. 2, the hall call allocation program 30 begins at the entry point 31 and the first subroutine 32 (or a series of subroutines) can use heuristic screening for upstream hall calls. It carries out against. This eliminates, for example, a car that is full and that call is before a call on the next floor of the car. As a result, the car is assigned due to a predetermined factor such as a car that has been assigned a “taxi” service corresponding to a call that has passed its duration, a car that is running in reverse depending on the location of the car in the building, or an up-peak assignment. Cars that are outside the reach of the car are excluded. All of these are conventional and are not part of the present invention.

  Step 33 sets the floor counter F to 1. Test 36 then determines whether there is an upward hall call HL CL on floor F. If not, the negative result of test 36 reaches step 37 and increments F, this time the next floor is tested to determine if there is a hall call on that floor.

  If there is an up hall call HL CL in the floor F, the step 38 shows the current waiting time CURR WAIT TIM for the up hall call HL CL in the floor F, and the time at which the floor F call was registered from the current clock time. Judge as subtracted. A subroutine (or series of subroutines) 39 then determines the remaining corresponding time RRT for each car C for all cars that can handle an up hall call at floor F. While this is preferably done according to the aforementioned US Pat. No. 5,672,853, the remaining corresponding time RRT may be determined in several other known ways.

  Subroutine 40, or a series of subroutines, then determines an expected new stop ENS for all cars C that can accommodate upstairs F calls. Next, in step 41, the car counter C is set to 1. Test 42 determines whether car C can handle an upstairs F call. If not, the C counter is incremented in step 43 and the process returns to the test 42 again. If the car is available, step 45 determines the expected waiting time EXPD WAIT, i.e., the estimated time until car C arrives at the upcall at floor F. This is the remaining response time RRT (determined in the subroutine 39) until the car C arrives at the upstairs call of the floor F, and the time CURR WAIT TIM (determined in step 38) that the passenger has waited until now. Is determined as the sum of In step 46, the sensory waiting time PWT is determined as the predetermined constant K1 multiplied by the square of the expected waiting time EXPD WAIT in step 45. However, functions other than the square may be used.

  In step 47, the predicted travel time EXP TRAV TIM multiplies the predetermined constant K2 by the distance between the predicted destination floor ESTD DESTN FLR and the floor F where the call is registered, and the predetermined constant K3 of the car C. Determined as the sum of the expected new stop ENS and the sum of the stop CMTD STPS established before arrival at floor F. The expected destination floor ESTD DESTN FLR may be the actual destination floor in the building that has the destination floor call button, or is determined from historical call dates and call data for calls made on floor F using artificial intelligence. It may be a different floor. Said artificial intelligence may or may not be aided by a process involving a neural network recently disclosed in US Pat. No. 5,672,853. Alternatively, the expected destination floor ESTD DESTN FLR may simply be a half of the distance between the floor F and the top floor of the building. Which is used is irrelevant to the present invention.

  In step 51, the sensory travel time (PTT) is determined as the square of the predicted travel time EXP TRAV TIM determined in step 47 multiplied by a constant K4. However, functions other than the square may be used. Next, in step 52, the sensory service time (PST), that is, the length of time that the passenger feels it took to reach the destination floor, PERCVD SERV TIM, is the sum of the sensory waiting time PWT and the sensory travel time PTT. Determined. Next, at step 53, the sensory service time PST is squared. Of course, steps 45 to 53 are not determined step by step as shown in the figure, but may be combined into one long calculation formula.

  The car assigned to the call that has been waiting for the longest arrival of the car is the car with the constant K1 and steps 46 and 51 to make it the car that delivers the passenger to its destination floor the fastest One or both of the functions may be adjusted with respect to the constants K2-K4 and the non-linear function (which is squared in step 51). In this way, the total sensory latency is reduced, including the time waiting for the elevator car to arrive and the time waiting for the car to deliver passengers to their destination floor.

  Test 54 determines whether all the cars that can service the incoming call at floor F have been tested. If not, the program returns to step 43 where C is incremented and the next car is tested in test 42 to determine if it can accommodate an upstairs F call. If yes, the process described above is repeated for that car. If not, the car count is incremented and the next car is tested.

  Once all the cars have been processed for an up call at floor F, test 55 determines whether all of the up floor calls have been processed. Since it is not initially, a negative result is obtained at test 55, which causes the routine to return to step 37 to increment F, and at test 36 to determine if the next floor has an upstream hall call. Is done. If so, the above process is repeated for this and subsequent floors until test 55 is positive, indicating that all upstream calls have been processed.

  A series of tests, steps and subroutines 56, similar to the tests, steps and subroutines 32-55, are then repeated for the downstream hall call.

  Another aspect of the invention is to use the system to strive to equalize sensory service time across all passengers. By using the least square sum of sensory service times for the car allocation to a call, a set of allocations with service times close to each other is selected. For example, assuming that Call A's PST is equal to 10 and Call B's PST is equal to 10, the sum of squares is 200. On the other hand, if call A's PST is equal to 9 and call B's PST is equal to 11, then the sum of squares is 202, which is the sum of the less desirable scenarios. If only the PST powers are summed, the result is 20 in both cases, and a more favorable total result is not shown. This is one embodiment of the present invention. Once processing of all uphole and downhole calls is complete, subroutine 59 obtains a square sum of sensory service times for all possible car assignments for up and down calls. This is done for all possible sets of assignments for all waiting uplink and downlink calls. Subroutine 60 then assigns cars based on a set of assignments that minimizes the square sum of sensory service times determined in subroutine 61.

FIG. 1 is a conceptual diagram showing a conventional computer configuration connected to an elevator as an example of a system in which the present invention can be implemented. FIG. 2 is a simplified logic flow diagram illustrating the process used to implement the present invention.

Claims (3)

A method of assigning a car selected from a plurality of elevator cars in order to deal with unsupported hall calls in a multi-story building ,
For serviceable each car with respect to unsupported specific hall call in buildings,
(A) determining a predicted value of the waiting time that elapses before the call, each of which is corresponding since their registration with the (45),
And (b) Rukoto give sensory latency as multiplied by the first non-linear function of the waiting time of each of the first constant (46),
(C) that the passenger calls the respective registers the call wherein each since the correspondence to determine a predicted value of travel time that elapses until it reaches the expected target floor and (47),
(D) and Rukoto give sensory travel time as obtained by multiplying the second non-linear function of the running time of each of the second constant (51),
For the waiting time and travel time corresponding to the (e), respectively, and Rukoto give total and to sensitive Satoshiteki service time of the sensory latency and the sensory travel time (52),
(F) allocating a serviceable cage so as to correspond to the unsupported hall call based on the sensory service time (59-61),
A method comprising the steps of: Hall call having a relatively long latency have you the car particular is to reach the expected destination floor in a relatively short running time in the particular car, the first constant and said first non-linear function The method according to claim 1, wherein the second constant and the second nonlinear function are selected together with the second constant . Allocating the car in (f) above
And Rukoto given square (53) of the sensory service time of each,
And non for the corresponding all possible allocations set for all the uplink hall calls and down hall calls given total of the squares (59) Rukoto in buildings,
According to a set of one having the smallest said sum of said set (61) to assign a car to the call (60),
The method according to claim 1, comprising a.

Images