CN105761037A - Logistics scheduling method based on space reverse neighbor search under cloud computing environment - Google Patents
Logistics scheduling method based on space reverse neighbor search under cloud computing environment Download PDFInfo
- Publication number
- CN105761037A CN105761037A CN201610083265.1A CN201610083265A CN105761037A CN 105761037 A CN105761037 A CN 105761037A CN 201610083265 A CN201610083265 A CN 201610083265A CN 105761037 A CN105761037 A CN 105761037A
- Authority
- CN
- China
- Prior art keywords
- space
- client
- data
- query
- logistics
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 238000000034 method Methods 0.000 title claims abstract description 66
- 238000012545 processing Methods 0.000 claims abstract description 11
- 238000010845 search algorithm Methods 0.000 claims abstract description 6
- 230000008569 process Effects 0.000 claims description 28
- 238000001914 filtration Methods 0.000 claims description 9
- 238000006062 fragmentation reaction Methods 0.000 claims description 8
- 238000013138 pruning Methods 0.000 claims description 8
- 238000013467 fragmentation Methods 0.000 claims description 6
- 238000012544 monitoring process Methods 0.000 claims description 5
- 238000011161 development Methods 0.000 claims description 4
- 230000005540 biological transmission Effects 0.000 claims description 3
- 239000003638 chemical reducing agent Substances 0.000 claims description 3
- 230000007246 mechanism Effects 0.000 claims description 3
- 238000013507 mapping Methods 0.000 claims description 2
- 230000001960 triggered effect Effects 0.000 claims description 2
- 238000005516 engineering process Methods 0.000 abstract description 6
- 230000009286 beneficial effect Effects 0.000 abstract description 2
- 230000002452 interceptive effect Effects 0.000 abstract 1
- HUTDUHSNJYTCAR-UHFFFAOYSA-N ancymidol Chemical compound C1=CC(OC)=CC=C1C(O)(C=1C=NC=NC=1)C1CC1 HUTDUHSNJYTCAR-UHFFFAOYSA-N 0.000 description 5
- 230000006870 function Effects 0.000 description 5
- 230000007812 deficiency Effects 0.000 description 4
- 238000013461 design Methods 0.000 description 4
- 238000000354 decomposition reaction Methods 0.000 description 3
- 239000012141 concentrate Substances 0.000 description 2
- 238000011160 research Methods 0.000 description 2
- 230000001502 supplementing effect Effects 0.000 description 2
- 230000004888 barrier function Effects 0.000 description 1
- 238000004364 calculation method Methods 0.000 description 1
- 239000007795 chemical reaction product Substances 0.000 description 1
- 238000004891 communication Methods 0.000 description 1
- 238000012790 confirmation Methods 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 230000006872 improvement Effects 0.000 description 1
- 230000006698 induction Effects 0.000 description 1
- 230000002045 lasting effect Effects 0.000 description 1
- 238000012423 maintenance Methods 0.000 description 1
- 230000035772 mutation Effects 0.000 description 1
- 238000005457 optimization Methods 0.000 description 1
- 230000008520 organization Effects 0.000 description 1
- 230000004044 response Effects 0.000 description 1
- 238000012546 transfer Methods 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/08—Logistics, e.g. warehousing, loading or distribution; Inventory or stock management
- G06Q10/083—Shipping
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/08—Logistics, e.g. warehousing, loading or distribution; Inventory or stock management
Landscapes
- Business, Economics & Management (AREA)
- Engineering & Computer Science (AREA)
- Economics (AREA)
- Quality & Reliability (AREA)
- Entrepreneurship & Innovation (AREA)
- Human Resources & Organizations (AREA)
- Marketing (AREA)
- Operations Research (AREA)
- Development Economics (AREA)
- Strategic Management (AREA)
- Tourism & Hospitality (AREA)
- Physics & Mathematics (AREA)
- General Business, Economics & Management (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention discloses a logistics scheduling method based on space reverse neighbor search under cloud computing environment, and belongs to the field of large-scale time-space data processing and mobile technology application. The location of a logistics truck is positioned and searched by adopting a large-scale distributed SRNN location-based service search algorithm. The big data processing mode is integrated to the service selection phase of the logistics scheduling system, and the logistics truck most meeting the conditions of a sending client is searched in mass data and information of the truck driver is fed back to a client side. Similarly, the truck driver can also position the sending client according to the request of the sending client and request or respond within a certain period of time. The beneficial effects of the logistics scheduling method are that the client location can be automatically matched, location precision is high, the method is suitable for the big data environment, an interactive platform provided for the sending client and the logistics truck driver can provide accurate logistics scheduling information, and object sending efficiency of the sending client can be enhanced.
Description
Technical field
The present invention relates to LBS (LocationBasedService) field, be a kind of Logistic Scheduling method based on the anti-NN Query in space under cloud computing environment, comprise the distributed process of extensive space-time data and the exploitation of intelligent mobile terminal application.
Background technology
Developing rapidly of computer and network technology thereof, the appearance of mobile intelligent terminal and the constantly ripe informationization for logistic industry of mobile positioning technique provide the basis of compacting.By mobile interchange, will be dispersed in the logistics van of diverse geographic location and outbox client is tightly linked and together, using logistics van, outbox client as node, is constituting the network of an information transmission and information sharing.Client can understand position and the functioning condition of neighbouring logistics van in real time, and logistics van driver can also receive the request of neighbouring client in real time accurately simultaneously, it is simple to the negotiation of each side's business, it is achieved that the efficient scheduling of logistics.
In conjunction with the fast development recently as location Based service and mobile Internet, in logistic dispatching system, the real-time query for node neighbor node each in network becomes a big technical barrier.It is on the one hand the swift and violent growth of the data volume due to geographical spatial data, on the other hand also in that for the particular/special requirement of real-time in Logistic Scheduling business.Therefore, how to realize efficient NN Query and become the new demand of the logistic dispatching system under mobile cloud computing environment and challenge.The method having had several NN Query now, such as most basic kNN method, and RNN (ReverseNearestNeighbor) the anti-NN Query as kNN problem mutation, have become as typical spatial query algorithms and obtain the common concern of industry.Since Korn et al. provides the definition of RNN inquiry first, as a hot research problem, there has been substantial amounts of achievement in research at present, and possessed good using value.
Due to existing most of RNN algorithm because the structurized feature of its index or algorithm, cause potential order executive problem.Do not possess extensibility it addition, have yet, lack precision, or produce the problems such as dimension disaster.It is simultaneously based on the rise of location-based service correlational study and application, stand-alone environment is because calculating limited with storage capacity, therefore extensive spacing query method cannot be supported, solution is exactly utilize many computing nodes to participate in calculating simultaneously, and this is accomplished by the new distributed index of design and parallel query method.Although RNN inquiry is equally important with other spacing query method, but effectively supports that the achievement of the extensive RNN algorithm aspect inquired about is relatively fewer.MRVoronoi, except the index structure of design space, also describes briefly and utilizes the MapReduce method carrying out RNN inquiry.But the method needs the extra time to position query point, may result in high maintenance and calculation cost when dimension increases.RankReduc method is supported to process extensive approximate kNN inquiry by MapReduce, but the method is coarse, is only used for higher-dimension condition, and does not describe the process solving RNN.So existing method all can not be used directly to the extensive RNN of solution well and inquire about problem.
Concentrate the problem making a look up RNN to serve not only as an Intelligent logistics dispensing Algorithm for Solving optimization problem from extensive spatial data, also have a wide range of applications in the fields such as intelligent navigation, traffic control, relief assistance, weather forecast, space clustering simultaneously.But in the face of magnanimity, large-scale space-time data, traditional RNN method, being difficult to the demand meeting mobile interchange logistics system for real-time, based on this starting point, we have designed and Implemented this invention.
Summary of the invention
According to the defect existed in above-mentioned background technology and deficiency, the invention provides the Logistic Scheduling method based on the anti-NN Query in space under cloud computing environment, to solve the deficiency that in existing logistics transfer system, outbox client calculates with logistics van near neighbor problem.The present invention has improved also for the deficiency in the anti-nearest Neighbor existed in prior art, in order to improve accuracy and real-time.
To achieve these goals, the technical solution adopted in the present invention is: based on the Logistic Scheduling method of the anti-NN Query in space under cloud computing environment, by cloud center service system and intelligent mobile FTP client FTP, performs step as follows:
S1. cloud center service system is for carrying out down the foundation of row's grid index, and performs distributed extensive anti-nearest neighbor algorithm;
S2. intelligent mobile client adapts to different user functional requirement and use habit, base station by built-in alignment system and operator, and Development by Depending on Network obtains self real-time spatial geographical locations, initiate inquiry request, and to carry out information mutual with cloud center service system.
Supplementing as technical scheme, the cloud center-side service system that this extensive anti-NN Query logistic dispatching system uses by no less than a Ge Yun data center the webserver or what fictitious host computer was constituted, adopt this parallelization of cloud computing calculating to process large-scale data and tackle in the substantial amounts of client needing to send out express delivery, and with client for Help Center, initiate inquiry to ask clearly in this position, logistics van is as the data set being available for inquiry, in such a mode, ensure that scheduling stability during high logistics capacity, accelerate response speed during logistics information search, enhance extensibility simultaneously.Cloud center service system, by using the Customer Location and logistics van position collected, is set up the distributed of positional information and is arranged grid index.
Used the row's of falling grid index to carry out the extensive anti-NN Query of distributed space time information by cloud center service system, and return optimum logistics van to outbox client.
The process step of the row's of falling grid index is particularly as follows: give space data sets P and Q, the set that P and Q is made up of Euclidean space data point, and data set P has Customer Location and Q has logistics van positional information, for the Customer Location point point q ∈ Q in P at the position of data set Q expression formula q (x, y) representing, some q comprises the positional information of logistic car and the customer information of periphery thereof.
First document data set is stored on distributed file system HDFS, HDFS can be divided into a lot of deblocking automatically, each Mapper reads in an input data fragmentation, then the spatial data points in each Mapper analytical data burst, and calculate the spatial data points mapping to grid cell lattice, last Mapper is cell p (i, j) position in the middle of grid is as key, some q (x, y) positional information is as value, and should<key, value>corresponding output, the data output that Reducer then reads Mapper, and collect the point data in same unit lattice (key), then the set of the point that output unit lattice index and are included in this cell.
Method based on the extensive anti-NN Query of grid index is: referring to Fig. 9, initially set up space lattice index, and mesh space is carried out entire scan, thus the row's of establishing grid index, to slice data area PCT wheel rim algorithm in Map function, with a ciFor the center of circle, radius r=| ci,si| carry out wheel rim, and the Counter (g of grid Cell (i) intersected in region or with round edge circle will be justifiedi) value is calculated as 1, i.e. Counter (gi)=1;After each slice data area individual processing is complete, finally merges in Reduce function, in the process of merging, be scanned according to Mesh Processing Algorithm, every time in the process of scanning, the Counter (g to overlapping grid Cell (i)i) value add up, finally export the whole area of space maximum Cell (j) of weights W.
Supplementing as technical scheme, extensive anti-NN Query is defined as: assume by a N dimension space D being made up of spatial object p and query object q, Reverse nearest neighbor inquiry is by traveling through all object p ∈ P and finding out satisfied: RNN (q)=and p ∈ P | Dist (p, q) < Dist (p, p') }.Here Dist () is the Euclidean distance between two objects, and p ' is the neighbour's object in P from p kth person.
Given Euclidean distance space data sets P (outbox client) and Q (logistics passenger vehicle), wherein P and Q is different types of data set, if 1 p in data-oriented collection P, SRNN Query Result is to return to be had a q ∈ Q, wherein q is the arest neighbors node of p, and p is also the Reverse nearest neighbor node of q simultaneously.
Referring to Fig. 1, if q is outbox client, look for nearest logistics van p.If it is simply simple with the nearest logistics van of kNN algorithm queries, then the addressee person p on logistics van3And p5Recommendation will be put into concentrate, but for addressee person p3, have from it closer to client, i.e. q2And q4, it is more willing to accept q2And q4Post part, then p3For q point be not suitable for recommended.Only when the p Reverse nearest neighbor node being also q simultaneously, just can be eventually served as SRNN Query Result and be returned.
Referring to Fig. 2, accurately and it is suitable for the extensive characteristic processing the existing some algorithm of three aspect induction and contrast and SRNN algorithm from various dimensions support, result.First, it is assumed that space data sets is magnanimity rank, the stand-alone server that internal memory and computing capability are limited cannot directly process;Secondly, algorithm is to run on the multidimensional metric space of 2-4 dimension, and in order to simplify problem, what computed range all adopted is Euclidean distance.3rd, the uniformly random distribution of data point right and wrong, it is generally configured with the data characteristics tilted;Finally, query point is not concentrated at initial data, is occur at random in data space, and all of data point and query point are invariant positions in timeslice, and temporally sheet is periodically subject to updating location information.The above is assumed all to be consistent with application demand with the data characteristics of most real world.
Beneficial effect: better solve the search inquiry between outbox client and logistics van driver and be mutually located problem, this Logistic Scheduling software has been studied in spatial data index and querying method Problems existing also for the deficiency in technology, a kind of method proposing improvement, improve the speed of the location under large-scale data environment and search inquiry, accuracy, degree of accuracy, and strengthen between truck man and outbox client mutual.
Accompanying drawing explanation
Fig. 1 be the RNN of the present invention set up process algorithm;
The RNN algorithm that Fig. 2 is the present invention is summed up;
Fig. 3 is SRNN filtration stage algorithm steps;
Fig. 4 is that Basic-SRNN example is shown;
Fig. 5 is scale logistics system Organization Chart;
Fig. 6 is the functional block diagram of the present invention;
Fig. 7 is the flow chart that outbox client and the courier of the present invention is mutual;
Fig. 8 is Pruning strategy example;
Fig. 9 is PCT wheel rim algorithm.
Detailed description of the invention
Embodiment 1: such as reference Fig. 5, based on the Logistic Scheduling method of the anti-NN Query in space under cloud computing environment, including cloud center service system and intelligent mobile FTP client FTP, wherein, cloud center service system arranges grid index for setting up the distributed of positional information, and extensive anti-NN Query (SRNN) algorithm between execution distributed space, intelligent mobile client is respectively for logistics van driver and outbox client.Outbox client includes the basic functions such as map, location, voice.Namely this system execution step is as follows:
S1. cloud center service system is for carrying out down the foundation of row's grid index, and performs distributed extensive anti-nearest neighbor algorithm;
S2. intelligent mobile client adapts to different user functional requirement and use habit, base station by built-in alignment system and operator, and Development by Depending on Network obtains self real-time spatial geographical locations, initiate inquiry request, and to carry out information mutual with cloud center service system.
Embodiment 2: there is the technical scheme identical with embodiment 1.Wherein realize the step included by Logistic Scheduling as follows: such as reference Fig. 6, after logistics van driver operationally signs in outbox client software, it is automatically positioned current position, then choose whether to disclose the position of oneself, if it is open, outbox client then can search its position, otherwise, then can not.After outbox client discloses the position of oneself, the positional information after oneself being changed uploads to cloud server, and data can be stored by cloud server, and automatically clears up expired data.Logistic Scheduling user is automatically positioned current position, first carries out the confirmation of destination before Logistic Scheduling after signing in outbox client mobile terminal software, it is determined that there are three kinds of modes in destination, is be manually entered, click map, voice typing respectively.Can start after determining to search for logistics van, cloud server uses spatial index algorithm that the data comprising logistics truck man position are processed according to the position of outbox client after receiving current request, find out the optimum logistics van that distance outbox Customer Location is nearest, and its information is returned to outbox client.
Embodiment 3: there is the technical scheme identical with embodiment 2.As shown in reference Fig. 7, wherein outbox client can show the logistics van searched on map, click logistics van icon and can check the specifying information of corresponding lorry, such as phone number etc., if outbox client wishes to be come by phone and truck man communication, the number in click information page then can directly invoke dialer and dial.After confirming Logistic Scheduling, truck man end then can receive the Logistic Scheduling information of correspondence, as truck man is agreed to then represent this Logistic Scheduling success, showing the position of both sides and a path between outbox client and logistics van on map, before outbox client waits logistics van, contact receives goods simultaneously.If truck man can not process Logistic Scheduling information in time, then it represents that concludes the business unsuccessful.
Embodiment 4: there is the technical scheme identical with embodiment 3, wherein: the execution method of this logistic dispatching system is: handheld device outbox client is by based on 2G/3G/4G mode or the wireless network of WIFI, set up with cloud server while accessing mobile Internet and contact, outbox client is responsible for display map, and carry relevant parameter, such as position data.Outbox customer information sends request to high in the clouds, logistics van driver is after logging in this grand scale logistic dispatching patcher, spatial geographic information (including oneself positional information and outbox customer information) is sent to cloud server, and lorry spatial geographic information disclosed in driver is set up distributed spatial index by spatial geographic information server beyond the clouds.The outbox client software information by cell phone map service acquisition present position, the positional information got is sent to cloud server, cloud server adopts SRNN inquiring technology, inquire apart from oneself nearest optimum logistics van rapidly from the outbox customer information of magnanimity, the truck man inquired occurs on the outbox client software map interface of outbox client, truck man on outbox customer selecting software interface, can send, to server, the mode asking or directly dialing truck man phone and send request, outbox client request is forwarded to the outbox client of truck man by server, after truck man receives transmission request, request is handled it and returns to server, it is sent to outbox Client handset outbox client by server.If truck man and outbox client reach an agreement, truck man will go to the place of outbox client to provide logistics service.
Embodiment 5: there is the technical scheme identical with embodiment 4, wherein: wherein SRNN query processing step is particularly as follows: SRNN query processing process, it is possible to be decomposed into two spatial manipulation processes independently: namely filter and Qualify Phase.(1) filtration stage: such as reference Fig. 3, the main purpose of filtration stage is to obtain the Candidate Set potentially including result by inquiring about.The core concept of filtration stage is to find all of neighbor objects around in the spatial dimension of q point.By utilizing design PCT algorithm to read the cell around q point concurrently, obtain initialized kNN query results to obtain the q all neighbour's objects concentrated at spatial data thus calculating.Specifically, first centered by q, initialize wheel rim radius rδ, afterwards the cell region of this round region overlay is monitored.At parallel increase radius r in algorithm performsδTime, new one can be triggered and take turns PCT algorithm and carry out incremental update neighbour object set P.So in iteration execution process subsequently, in unit interval sheet, it is only necessary to monitoring radius rδCorresponding region S is without monitoring whole data space.Execution PCT algorithm eventually through increment can find all neighbours of q surrounding space and as candidate result collection Scnd.In the calculating process of whole filtration stage, it is that PCT is performed as a kind of expansible search algorithm that large-scale spatial object can be supported under distributed environment to inquire about.(2) Qualify Phase: Qualify Phase checks all of kNN Candidate Set mainly by calculating, gets rid of the point being not belonging to RNN in Candidate Set thus obtaining finally correct RNN result.Specifically, in filtration stage above, the Candidate Set S obtainedcnd.Then, ScndIn each point as query point, by performing distributed search algorithm kNN, calculate ScndEach is concentrated to put respective reverse neighbours to verify that whether it is the RNN of q.For each Candidate Set ScndIn point, adopt distributed mechanism to be concurrently verified.In each parallel task, between falling to empty by lasting traversal, grid index execution kNN inquiry is verified, until the point in all of Candidate Set has all been verified, and by finally correct RNN result output.
Embodiment 6: there is the technical scheme identical with embodiment 4 or 5, wherein SRNN query steps according to distributed system thought of dividing and rule can be divided into decomposition (Divide) and two subprocess that merge (Conquer).In decomposing (Divide) function, process filtration stage obtain Candidate Set, then by carrying out result merger and verifying that the point of debug obtains final RNN result set in merging (Conquer) function.Detailed process is: (1) decomposition step: a given Candidate Set ScndSome pi(i=1,2 ..., k).As shown in reference Fig. 4, filtration stage obtains the Candidate Set S of qcndFor { p1,p2,p3,p4,p5,p6}.In decomposition step, first Candidate Set ScndIn point as query point, run PCT method and carry out the cell around this point of rounds of readings, thus obtaining ScndIn the arest neighbors of point.Using all neighbours of q of inquiring as Candidate Set, data space is divided into several file fragmentations, is divided into four file fragmentations Split1, Split2, Split3 and Split4 with reference to Fig. 4.Mapper reads output { (p after file fragmentation Split3 processes1,q),(p2,p1), read output { (p after split4 processes4,p5),(p6,p5)}.If data need not process for empty (not having data point) in burst, if the arest neighbors of a data point is not in same burst, query processing respectively will be carried out across burst, and in merging process, finally carry out result collect process.(2) combining step: Reducer receives query point as Key, kNN candidate point as Value from the output result set of each Mapper, carries out result merging by same Key.Then being calculated by distance in this step and go to verify relative to other object-point, whether q is and piNearest point, if not words be not just correct result and be excluded, the p in reference Fig. 41And p3Arest neighbors be all q.And < p4,p5> arest neighbors each other, do not comprise q, so being excluded, in like manner other point is all excluded.Therefore, { p1,p3For the end product collection of SRNN (q).
In the process that MapReduce framework performs, when Map task terminates, passing through process of shuffling, this just has substantial amounts of data from local Map node by being transferred to Reduce node.The substantial amounts of data shuffled in process in this Map Reduce system move and can cause serious system load, if it is possible in the intermediate conveyor data of the process of shuffling after minimizing Map task, it is possible to significantly increase the treatment effeciency of MapReduce.So needing to design some more effective RNN to filter search algorithm and optimisation strategy.Unnecessary intermediate result set is cut, it is possible to be effectively improved the whole efficiency utilizing Map Reduce system to perform RNN inquiry by two points of hyperplane pruning algorithms of distributed Pruning strategy.It is given below according to two points of hyperplane pruning algorithms, carries out an example of beta pruning, in reference Fig. 8, figure, provide query point q (delivery client) and some object-point p1To p7(delivery vehicle), in the first wheel rim, remaining area is initialized as overall data space, it is possible to the arest neighbors Candidate Set obtaining q is { p1,p4,p7}.At q and p1Vertical two minutes red dotted line ⊥ (p are produced between point1, q) (because the independence of data fragmentation, so this vertical equinoctial line, only produce in the space at this data fragmentation place), at equinoctial line ⊥ (p1, the q) p in region, upper right side2Point, it is possible to learn Dist (p2,p1)<Dist(p2, q), so p2Point can be fallen by beta pruning and be not involved in wheel rim below and calculate.In like manner, it is possible to obtain ⊥ (p4, q) with ⊥ (p7,q).MapReduce framework has been improved by we, Master adds Distributed Cache Mechanism, its state monitoring all of Mapper can be passed through, so can be obtained by the cell in remaining region (being designated the cell of gray shade), and proceed the second wheel rim traversal wherein, find the neighbour p of q3, now there is no any data point in non-beta pruning cell, so final Candidate Set is Scnd={ p1,p4,p7,p3}。
The above; it is only the present invention preferably detailed description of the invention; but protection scope of the present invention is not limited thereto; any those familiar with the art is in the technical scope of present disclosure; it is equal to replacement according to technical scheme and inventive concept thereof or is changed, all should be encompassed within protection scope of the present invention.
Claims (6)
1. based on the Logistic Scheduling method of the anti-NN Query in space under a cloud computing environment, it is characterised in that: by cloud center service system and intelligent mobile FTP client FTP, perform step as follows:
S1. cloud center service system is for carrying out the foundation of the distributed row of falling grid index, and performs distributed extensive anti-nearest neighbor algorithm;
S2. intelligent mobile client adapts to different user functional requirement and use habit, base station by built-in alignment system and operator, and Development by Depending on Network obtains self real-time spatial geographical locations, initiate inquiry request, and to carry out information mutual with cloud center service system.
2. based on the Logistic Scheduling method of the anti-NN Query in space under cloud computing environment according to claim 1, it is characterized in that: after outbox client sends transmission logistics express delivery association requests, automatically gathered and submitted to the positional information of logistics van near outbox client and outbox client to cloud center service system by the alignment system of intelligent mobile client, cloud center service system is by the positional information of the logistics van near the outbox client collected and outbox client, off-line is set up the distributed of outbox Customer Location information and is arranged grid index, and the position of outbox Customer Location and neighbouring logistics van thereof is carried out distribution pretreatment, and carry out as required regularly dynamically updating;Used the row's of falling grid index to carry out the Reverse nearest neighbor inquiry of distributed space time information by cloud center service system, and return optimum logistics van information to outbox client.
null3. based on the Logistic Scheduling method of the anti-NN Query in space under cloud computing environment according to claim 2,It is characterized in that: the process step of the distributed row of falling grid index is particularly as follows: give space data sets P and Q,The set that P and Q is made up of Euclidean space data point,Data set P has Customer Location and Q has logistics van positional information,For the Customer Location point p in P,Inquire about the q ∈ Q position expression formula q (x at data set Q,Y) represent,Point q comprises the positional information of logistic car and the customer information of periphery thereof,Above large data sets file is stored on distributed file system HDFS,HDFS can be divided into a lot of deblocking automatically,Each Mapper reads in an input data fragmentation,Then each Mapper reads the spatial data points in data fragmentation,And calculate the mapping to grid cell lattice of each spatial data points,Last Mapper is cell p (i,J) position in the middle of grid is as key,Some q (x,Y) positional information is as value,And should < key,Value > corresponding output,Reducer then reads the data output of Mapper,And collect and the point data in same unit lattice (key) is carried out shuffle operation,Then the set of the point that output unit lattice index and are included in this cell.
4. based on the Logistic Scheduling method of the anti-NN Query in space under the cloud computing environment according to any one of claim 1-3, it is characterized in that: extensive anti-NN Query (SRNN) is defined as: assume by a N dimension space D being made up of spatial object p and query object q, Reverse nearest neighbor inquiry is by traveling through all object p ∈ P and finding out satisfied: RNN (q)=and p ∈ P | Dist (p, q) < Dist (p, p') }, here Dist () is the Euclidean distance between two objects, p' is from the neighbour's object away from p kth in P, SRNN query processing process, two spatial manipulation processes independently can be decomposed into: wherein filtration stage is: centered by p, initialize wheel rim radius rδ, the cell region of this round region overlay is monitored, when increasing radius r in executed in parallelδTime, new one can be triggered and take turns PCT in order to incremental update neighbour's object, perform process through iteration, in unit interval sheet, it is only necessary to monitoring radius rδCorresponding region S is without monitoring whole data space, and the execution PCT algorithm eventually through increment can find all neighbours of p surrounding space and as candidate result collection Scnd, whole calculating process, is that PCT is performed as a kind of expansible search algorithm that large-scale spatial object can be supported under distributed environment to inquire about;Qualify Phase: Qualify Phase checks all of S mainly by calculatingcndCandidate Set, gets rid of the point being not belonging to RNN in Candidate Set thus obtaining finally correct RNN result, ScndIn each point as query point, by performing distributed search algorithm kNN, calculate ScndEach is concentrated to put respective reverse neighbours to verify that whether it is the RNN of q, for each Candidate Set ScndIn point, adopt distributed mechanism to be concurrently verified.
5. based on the Logistic Scheduling method of the anti-NN Query in space under cloud computing environment according to claim 4, it is characterised in that: the application of two points of hyperplane pruning methods in extensive Reverse nearest neighbor inquiry.
6. based on the Logistic Scheduling method of the anti-NN Query in space under cloud computing environment according to claim 5, it is characterized in that: the method inquired about based on the extensive Reverse nearest neighbor of grid index is: initially set up space lattice index, and mesh space is carried out entire scan, thus the row's of establishing grid index, to slice data area PCT wheel rim algorithm in Map function, with a ciFor the center of circle, radius r=| ci,si| carry out wheel rim, and the Counter (g of grid Cell (i) intersected in region or with round edge circle will be justifiedi) value is calculated as 1, i.e. Counter (gi)=1;After each slice data area individual processing is complete, finally merges in Reduce function, in the process of merging, be scanned according to Mesh Processing Algorithm, every time in the process of scanning, the Counter (g to overlapping grid Cell (i)i) value add up, finally export the whole area of space maximum Cell (j) of weights W.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201610083265.1A CN105761037A (en) | 2016-02-05 | 2016-02-05 | Logistics scheduling method based on space reverse neighbor search under cloud computing environment |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201610083265.1A CN105761037A (en) | 2016-02-05 | 2016-02-05 | Logistics scheduling method based on space reverse neighbor search under cloud computing environment |
Publications (1)
Publication Number | Publication Date |
---|---|
CN105761037A true CN105761037A (en) | 2016-07-13 |
Family
ID=56330693
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201610083265.1A Pending CN105761037A (en) | 2016-02-05 | 2016-02-05 | Logistics scheduling method based on space reverse neighbor search under cloud computing environment |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN105761037A (en) |
Cited By (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN106599188A (en) * | 2016-12-14 | 2017-04-26 | 大连交通大学 | Smart store location method employing sub-space Skyline query under mobile internet and cloud computing environment |
CN106777093A (en) * | 2016-12-14 | 2017-05-31 | 大连大学 | Skyline inquiry systems based on space time series data stream application |
CN108197874A (en) * | 2018-01-23 | 2018-06-22 | 余绍志 | A kind of logistic management system based on Internet of Things big data |
WO2020206665A1 (en) * | 2019-04-12 | 2020-10-15 | Grabtaxi Holdings Pte. Ltd. | Distributed in‐memory spatial data store for k‐nearest neighbour search |
Citations (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN102214215A (en) * | 2011-06-07 | 2011-10-12 | 陆嘉恒 | Rapid reverse nearest neighbour search method based on text information |
CN103488679A (en) * | 2013-08-14 | 2014-01-01 | 大连大学 | Inverted grid index-based car-sharing system under mobile cloud computing environment |
CN105183921A (en) * | 2015-10-23 | 2015-12-23 | 大连大学 | Shop addressing system based on bi-chromatic reverse nearest neighbor inquiry under mobile cloud computing environment |
-
2016
- 2016-02-05 CN CN201610083265.1A patent/CN105761037A/en active Pending
Patent Citations (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN102214215A (en) * | 2011-06-07 | 2011-10-12 | 陆嘉恒 | Rapid reverse nearest neighbour search method based on text information |
CN103488679A (en) * | 2013-08-14 | 2014-01-01 | 大连大学 | Inverted grid index-based car-sharing system under mobile cloud computing environment |
CN105183921A (en) * | 2015-10-23 | 2015-12-23 | 大连大学 | Shop addressing system based on bi-chromatic reverse nearest neighbor inquiry under mobile cloud computing environment |
Non-Patent Citations (1)
Title |
---|
季长清: "云计算环境下的大规模空间近邻查询算法研究", 《中国博士学位论文全文数据库 信息科技辑》 * |
Cited By (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN106599188A (en) * | 2016-12-14 | 2017-04-26 | 大连交通大学 | Smart store location method employing sub-space Skyline query under mobile internet and cloud computing environment |
CN106777093A (en) * | 2016-12-14 | 2017-05-31 | 大连大学 | Skyline inquiry systems based on space time series data stream application |
CN108197874A (en) * | 2018-01-23 | 2018-06-22 | 余绍志 | A kind of logistic management system based on Internet of Things big data |
WO2020206665A1 (en) * | 2019-04-12 | 2020-10-15 | Grabtaxi Holdings Pte. Ltd. | Distributed in‐memory spatial data store for k‐nearest neighbour search |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN105760470A (en) | Medical calling system based on spatial reverse nearest neighbor query in cloud computing environment | |
CN105760465A (en) | Medical calling method based on large-scale reverse nearest neighbor query in mobile environment | |
Silva et al. | Towards sustainable smart cities: A review of trends, architectures, components, and open challenges in smart cities | |
Held et al. | The traveling-salesman problem and minimum spanning trees | |
Xu et al. | Cloud-based fleet management for prefabrication transportation | |
CN105183921B (en) | The shop addressing system based on double-color reverse NN Query under mobile cloud computing environment | |
CN103488679A (en) | Inverted grid index-based car-sharing system under mobile cloud computing environment | |
CN103884345B (en) | Interest point information collecting method, interest point information displaying method, interest point information collecting device, interest point information displaying device, and interest point retrieval system | |
CN105761037A (en) | Logistics scheduling method based on space reverse neighbor search under cloud computing environment | |
Zhang et al. | Vehicle-based bi-objective crowdsourcing | |
CN103559636A (en) | Restaurant management system based on cloud computing and intelligent analysis | |
CN101370025A (en) | Storing method, scheduling method and management system for geographic information data | |
CN104240163A (en) | Intelligent district recommendation method and intelligent district recommendation device | |
CN101488158A (en) | Road network modeling method based on road element | |
CN107146026B (en) | Space planning land difference automatic coordination method | |
CN105787585A (en) | Logistics scheduling system based on large-scale reverse nearest neighbor (SRNN) query in mobile environment | |
Li et al. | Two-sided online bipartite matching in spatial data: experiments and analysis | |
CN107658010A (en) | Portable medical querying method and application based on the row's of falling Thiessen polygon index | |
CN107291963B (en) | KNN query method and system under road network moving environment | |
Vahidnia et al. | Heuristic game-theoretic equilibrium establishment with application to task distribution among agents in spatial networks | |
Leung et al. | Community logistics: a dynamic strategy for facilitating immediate parcel delivery to smart lockers | |
CN116562738A (en) | Intelligent freight dispatching method, device, equipment and storage medium | |
Li et al. | Geo-DMP: A DTN-Based Mobile Prototype for Geospatial Data Retrieval | |
CN102012908B (en) | Method for inquiring visible moving neighbours objects in environment with barriers | |
Cui et al. | Place your next branch with mile-run: Min-dist location selection over user movement |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
C06 | Publication | ||
PB01 | Publication | ||
C10 | Entry into substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
RJ01 | Rejection of invention patent application after publication |
Application publication date: 20160713 |
|
RJ01 | Rejection of invention patent application after publication |