CN101504795B - Working method for DSP control system applied to multi-storied garage parking position scheduling - Google Patents
Working method for DSP control system applied to multi-storied garage parking position scheduling Download PDFInfo
- Publication number
- CN101504795B CN101504795B CN2008101528029A CN200810152802A CN101504795B CN 101504795 B CN101504795 B CN 101504795B CN 2008101528029 A CN2008101528029 A CN 2008101528029A CN 200810152802 A CN200810152802 A CN 200810152802A CN 101504795 B CN101504795 B CN 101504795B
- Authority
- CN
- China
- Prior art keywords
- algorithm
- ant
- control system
- dsp control
- individuality
- 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.)
- Expired - Fee Related
Links
Images
Landscapes
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention relates to a method for operating a DSP control system applied to parking stall dispatching of a three-dimensional garage, which is characterized by comprising the following steps: firstly, confirming that the garage is in the operating state; secondly, storing vehicle signals; thirdly, beginning information processing by a DSP control system; fourthly, encoding parameters; fifthly, setting an initial group; sixthly, designing a fitness function; seventhly, setting a control parameter; eighthly, solving a GAAA algorithm and a vehicle collection and storage optimization path; and ninthly, switching a control language and outputting a control signal. The method has the advantages that: firstly, the GAAA algorithm is superior to the conventional method for only automatically and randomly storing vehicles into the garage and taking the vehicles out of the garage by utilizing sensing and photoelectric elements, overcomes the respective defects of the genetic algorithm and the ant colony algorithm, and is a good heuristic algorithm both in time efficiency and solving efficiency; secondly, the method adopts the DSP control system, has simple hardware circuit, safe and reliable output of trigger pulses and high real-time control precision, and can greatly improve the stability and the reliability of devices; and thirdly, the method has wide market application prospect.
Description
(1) technical field:
The present invention relates to a kind of method of work of DSP control system, particularly a kind of method of work that is applied to the DSP control system of multi-storied garage parking position scheduling.
(2) background technology:
At present, there is following problem mostly in parking position scheduling control system in conventional stereo garage:
The one,, the conventional stereo garage just utilizes sensing and photovalve that the scheduling of automated randomized ground access is carried out in the parking stall, and that the management of access sequence and parking stall does not also reach is scientific, intelligent, rationalize;
The 2nd,, though genetic algorithm has the quick global search capability, do not utilize for the feedback information in the system, often cause the redundant iteration of doing nothing, it is low to find the solution efficient;
The 3rd,, ant group algorithm is to converge on optimal path by the accumulation of pheromones and renewal, have distribution, parallel, global convergence ability, but initial stage pheromones scarcity causes algorithm speed slow;
The 4th,, the multi-storied garage action is complicated, require control system to realize sequentially-operating control, speed control, positioning control and safety interlock control, because the advantage of programmable controller (PLC) can both adapt to and satisfy the high performance request for utilization of multi-storied garage, therefore all adopt the core of PLC at present nearly all automatic parking system as three-dimensional garage control system, but because PLC control system I/O counts and the restriction of arithmetic speed, PLC but exists the problem of aspects such as travelling speed and production cost when being applied to multi-storied garage parking position scheduling, have influence on the whole efficiency and the utilization factor in garage, cause the waste of vehicle on access time and access car energy consumption.
(3) summary of the invention:
The object of the present invention is to provide a kind of method of work that is applied to the DSP control system of multi-storied garage parking position scheduling, this method is based on GAAA algorithm (Genetic And Ant Algorithm), it be with genetic algorithm (Genetic Algorithm) and ant group algorithm (Ant Algorithm) mutually fusion application in the DSP of multi-storied garage parking position scheduling control system, it makes full use of the senior decision making function of computing machine, and utilize the digital signal processor spare (DSP) that has become current new hot spot technology to do real-time processing, realization is pushed the automated three-dimensional garage to the intelligent development direction to the optimal control of multi-storied garage parking position scheduling.
Technical scheme of the present invention: a kind of method of work that is applied to the DSP control system of multi-storied garage parking position scheduling, its working environment is by exchanging traversing motor at one, lifting motor, frequency converter, sensor, various mechanical switchs, wire rope, the parking stall framework, in the multi-storied garage that conventional equipment such as vehicle-containing and traversing guide rail constitutes, it is characterized in that: it is a senior decision making function of utilizing computing machine, genetic algorithm (Genetic Algorithm) and ant group algorithm (Ant Algorithm) combined constitute algorithm based on GAAA, be used for the method for multi-storied garage parking position scheduling by the DSP control system, it comprises following job step:
(1) confirm that the garage duty is ready: confirm earlier whether the garage is ready, can whether each equipment all can safe and reliable work in the garage, finish access car task with normal operating conditions;
(2) access car signal: receive access car signal in the garage, detect the garage internal memory state of picking up the car, when control system receives user's bicycle parking or during the signal of picking up the car, the DSP control system confirms that by the scanning to database this signal is bicycle parking or picks up the car;
(3) DSP control system start information is handled: the DSP control system begins to carry out information processing according to the GAAA algorithm of genetic algorithm and ant group algorithm fusion;
(4) parameter coding: at first, the direction of each parking stall action is carried out parameter coding, and utilize the rand function to generate the decimal system real coding population of some at random;
(5) setting of initial population: to colony's operation of evolving, setting the initial population data that some represent initial search point by genetic algorithm, is that the next stage output of optimal path code generates initial initial population;
(6) design of fitness function: promptly to the evaluation of hereditary ideal adaptation degree, be meant in genetic algorithm, determine that with the size of ideal adaptation degree this individuality is genetic to the probability in the colony of future generation; By the ability assessment to hereditary ideal adaptation environment, as the foundation of selection operation, it is to be formed by the objective function conversion; After fitness value function by design came out each chromosomal fitness value calculation, the individual inheritance of holding according to " survival of the fittest " principle was to the next generation, and bad individuality is eliminated; Then, OX crossover operator, the conversion mutation operator according to modified form in the genetic algorithm carries out calculated crosswise and conversion variation to it;
(7) setting of controlled variable: determine to the far-reaching parameter value of performance of genetic algorithms;
(8) asking for of GAAA algorithm and access car optimal route: the maximum-minimum ant MMAS of system algorithm is designed and improves, in the GAAA algorithm, ant algorithm adopts the maximum-minimum ant MMAS of system algorithm, and utilize the initial population data that configure in the step (5) to carry out the recursive iteration of algebraically, to producing some groups of optimization solutions after the recursive iteration, adopt the maximum-minimum ant MMAS of system algorithm that these several groups of final outputs of optimization solution recursive iteration are preferably separated, promptly obtain the shortest path of vehicle access;
(9) conversion and control language and output control signal: convert the shortest path of the access vehicle obtained in the step (8) to control language, by DSP control system output signal, facilities such as control garage motor, frequency converter, vehicle-containing are finished the access car.
The direction of parking stall action is total in the above-mentioned said step (4): static, motionless, upwards, five kinds of states downwards, left and to the right; The coding method of adopting is: the odd bits of each individuality (being each chromosome in the above-mentioned said step (6)) is represented the car item, even bit is represented direction, respectively with 0,1,2,3,4 transfixions of representing the parking stall, upwards, this five kinds of states downwards, left and to the right; Form is as follows:
9 1 20 2 5 2 □□
The action of car item action car item action car item
Two kinds of methods are generally arranged: 1. according to the intrinsic knowledge of problem, manage to hold the distribution range of optimum solution take up space in whole problem space, set initial population in the setting of above-mentioned said step (5) initial population; 2. generate earlier the individuality of some at random, therefrom choose best individuality and be added in the initial population; The continuous iteration of this process is up to reaching predetermined scale; And through producing some groups of optimization solutions after the recursive iteration of setting algebraically.
The fitness of the individuality of the design of the middle fitness function of above-mentioned said step (6) is genetic to follow-on probability with this individuality and is directly proportional; The unique requirement of fitness function is that its result is a nonnegative value; The change of scale of fitness is certain mapping transformation to the objective function codomain, can overcome the prematurity convergence and roam phenomenon at random.
The individual inheritance of holding according to " survival of the fittest " principle in the above-mentioned said step (6) is divided into direct heredity or produces the heredity again of new individuality by intersecting to of future generation.
Design to the fitness function of genetic algorithm in the above-mentioned said step (6) comprises the design of selecting operator, crossover operator, mutation operator: 1. select operator: select operator also to be called and duplicate operator, be from contemporary community, to select some better individualities, and it is copied in the colony of future generation, the individual probability that duplicates is proportional to its adaptive value; 2. crossover operator: crossover operator is meant to be replaced the part-structure of two parent individualities reorganization and is generated new individual operation, the individuality that combination makes new advances, at strand space efficient search is arranged, reduce the failure probability to effective model simultaneously, various crossover operators all comprise two substances: determine the position of point of crossing and carry out the exchange of portion gene; 3. mutation operator: mutation operator is meant to be replaced some genic values in the individual coded strings with other genic value, form a new individuality, and the variation algorithm comprises the position of definitive variation point and carries out the genic value replacement.
Controlled variable mainly comprises the long l of string, group size n, crossover probability p in the above-mentioned said step (7)
c, the variation Probability p
m, wherein, the general parameter area of determining is: n=20～200, p
c=0.5～1.0, p
m=0～0.05, the long l that goes here and there sets according to actual conditions.
Ant algorithm in the above-mentioned said step (8) is based on bionics, promptly calculate ability that ant relies on optimum path search and can find shortest path between ant nest and the food, comprise ant stay the disappearance of As time goes on can intensity volatilizing gradually of a kind of volatility secretion (pheromone) that is called pheromones and pheromones on the path of process; Ant can this material of perception in the process of looking for food existence and intensity thereof, and instruct the direction of motion of oneself with this, tend to move, promptly select the intensity of this material on the probability in this path and this paths at that time to be directly proportional towards the high direction of this material intensity; The path that pheromones intensity is high more selects its ant just many more, and then the intensity of the pheromones that stays on this path is just bigger, and the big pheromones of intensity attracts more ant, thereby form a kind of positive feedback,, make most ant all can walk this optimal path by this positive feedback.
Maximum in the above-mentioned said step (the 8)-minimum ant MMAS of system algorithm is a kind of based on bionic method, promptly calculate ability that ant relies on optimum path search and can find shortest path between ant nest and the food, comprise ant stay the disappearance of As time goes on can intensity volatilizing gradually of a kind of volatility secretion (pheromone) that is called pheromones and pheromones on the path of process; Ant can this material of perception in the process of looking for food existence and intensity thereof, and instruct the direction of motion of oneself with this, tend to move, promptly select the intensity of this material on the probability in this path and this paths at that time to be directly proportional towards the high direction of this material intensity; The path that pheromones intensity is high more selects its ant just many more, and then the intensity of the pheromones that stays on this path is just bigger, and the big pheromones of intensity attracts more ant, thereby form a kind of positive feedback,, make most ant all can walk this optimal path by this positive feedback.
Maximum-minimum ant the MMAS of system algorithm has three aspects different with the standard ant group algorithm in the above-mentioned said step (8):
1. pheromones update mode difference, each circulation back is only carried out pheromones to this circulation optimum solution or an ant up to the present finding out optimum solution and is upgraded, and in the standard ant group algorithm, all ants path of passing by is all carried out pheromones and upgraded, MMAS pheromones update mode is as follows:
τ
ij(t+n)＝ρτ
ij(t)+Δτ
ij best，Δτ
ij best＝1/f(s
best) (i)
τ wherein
Ij(t) expression t stays path ij constantly and goes up pheromones track amount, and ρ is a coefficient, Δ τ
Ij BestRepresent that best ant stay quantity of information on the ij of path, f (s in this circulation
Best) expression iteration optimum solution (s
Ib) or globally optimal solution (s
Gb) value;
2. the stagnation for avoiding searching for, the codomain scope of the pheromones track amount on each element of separating is limited in [τ
Min, τ
Max] in the interval.That is: if τ is arranged
Ij(t)＞τ
Max, then put τ
Ij(t)=τ
MaxIf τ is arranged
Ij(t)＜τ
Min, then put τ
Ij(t)=τ
MinAnd pheromones track amount is not limited in standard ant system, make track amount on some paths far above other limits, thereby ant has stoped the more excellent behavior of separating of further search all along moving with paths;
3. for to make ant can search for new solution more, the pheromones Track Initiation is turned to τ in the starting stage of algorithm
Max, MMAS is owing to be limited in [τ to the codomain of the pheromones track amount on each path
Min, τ
Max] in the interval.
In the above-mentioned said step (8) maximum-minimum ant MMAS of system algorithm is designed and improves, wherein the pheromones update mode to MMAS is: the first, and because of pheromones on each path among the MMAS is initialized as τ
Max, should reduce the pheromones on the relatively poor path rapidly, accelerate search speed, so after each circulation, calculate the mean value L of each paths length earlier
AveThen, increase less than the routing information of median length is plain, routing information element greater than median length reduces, second, the amount of the increase of each side information element or minimizing should be the same on each paths, and the length on each limit that should embody each paths is to the percentage contribution of path total length, so (i) formula is modified as following form:
τ
ij(t+n)＝ρτ
ij(t)+Δτ
ij，Δτ
ij＝∑Δτ
ij k (ii)
Δ τ wherein
Ij kRepresent that k ant stay quantity of information on the ij of path, Δ τ in this circulation
IjRepresent to stay in this circulation the informational capacity on the ij of path, wherein Δ τ
Ij kAs follows:
L wherein
kIt is the path that k ant passed by in this circulation; L
AveAverage length for m path that ant walks; d
IjBe the distance of city i to city j.
Principle of work of the present invention is: the DSP control system is the advantage that has more that proposes with respect to traditional PLC control system and the control system of development prospect; The problem of aspects such as travelling speed that in the application of multi-storied garage, is occurred for conventional P LC control system and production cost, the DSP control system has than PLC control system more superiority: at first aspect the travelling speed, the arithmetic speed of PLC control system can only reach the us level, and the arithmetic speed of DSP control system can arrive the ns level, therefore can satisfy the requirement of electronics high-intelligentization multi-storied garage for control chip better; Aspect production cost, because the I/O that multi-storied garage needs counts often, if use the PLC control system, production cost will be very high, but can reduce greatly on production cost with the words that DSP+CPLD realizes, the method that present DSP control system extends out CPLD has all obtained using widely in military affairs, industrial High Accuracy Control, digital product; Aspect the function expansion, the DSP control system has become various industrial various modules commonly used at chip integration, for the upgrading of multi-storied garage function provides powerful hardware guarantee; As can be seen above-mentioned, the DSP control system extends out CPLD can replace the main flow that the PLC control system becomes hyundai electronics smart three-dimensional garage control section gradually.
Parameter coding is the coding/decoding method when using genetic algorithm and solving individual chromosome spread pattern or the individual phenotype that changes to solution space from the search volume to the genotype.
The subject matter of garage operation is to try to achieve the shortest path of garage operation under the restricted constraint condition of parking stall moving direction; At first the direction of its action is carried out parameter coding, the action of parking stall is always total: static, motionless, upwards, 5 kinds of states downwards, left and to the right; Thus, can adopt such coding method: the odd bits of each individuality (chromosome) is represented the car item, even bit is represented direction, respectively with 0,1,2,3,4 transfixions of representing the parking stall, upwards, this five kinds of states downwards, left and to the right, and utilize the rand function to generate the decimal system real coding population of some at random, for convenience fitness function relatively sort and calculate on this basis select probability, the value of fitness function to get on the occasion of; Usually, we take minimum value to be converted into peaked method, promptly deduct the method for objective function with the constant that bigger and colony is irrelevant; For the present invention, the action of the even number element representation vehicle-containing of every row with the number addition of non-zero, has just obtained the number of times of whole garage vehicle-containing action; After each chromosomal fitness value calculation come out, next step work is exactly will be according to " survival of the fittest " thought, the individuality of holding is genetic directly to the next generation or produces new individuality by intersection and is genetic to the next generation again, and bad individuality is eliminated, this step operation is called selection; The method that the present invention has adopted roulette wheel to select, it can reflect ideal adaptation degree shared ratio in the ideal adaptation degree summation of whole colony preferably.That is to say that individual each selected probability is directly proportional with its relative adaptation degree in group environment; Traditional hereditary crossover operator does not fully take into account the characteristics of multi-storied garage parking position scheduling, the premium properties on the round can not be genetic in the colony of future generation well, and optimizing speed is relatively poor; The present invention has adopted a kind of OX crossover operator of modified form, has solved this problem preferably.In order to safeguard the diversity of colony, prevent the premature convergence of genetic algorithm, this paper has introduced the conversion mutation operator, also be the counter-rotating mutation operator, kept good gene segment, enable to entail the next generation, the while has produced again and has comprised complicated more new individuality, has enlarged the hunting zone effectively; During the operation of this operator, in the path, select point of penetration at random, then with the sub-road counter-rotating between these 2; According to genetic operator design of the present invention, population scale is 20, and iterations was 20 generations, crossover probability p
c=0.8, the variation Probability p
m=0.2, some groups of optimization solutions that produced by genetic algorithm GA, adopt above-mentioned said maximum-minimum ant MMAS of system that the shortest path of also promptly obtaining vehicle access is preferably separated in these several groups of final outputs of optimization solution recursive iteration, and then finish the access car by facilities such as DSP control system control garage motor, frequency converter, vehicle-containings.
Superiority of the present invention is: 1. the GAAA algorithm at first utilizes random search, rapidity, the global convergence of genetic algorithm to produce plain distribution of initial information of relevant issues.Then, make full use of ant group's concurrency, positive feedback mechanism and find the solution features such as efficient height, not only be better than traditional just utilize sensing and photovalve and vehicle deposited automated randomizedly and takes out in the parking stall in garage, and genetic algorithm and ant group algorithm shortcoming have separately been overcome, algorithm after merging like this all is reasonable heuritic approach in time efficiency with finding the solution on the efficient; 2. adopt the DSP control system to replace original PLC control system, hardware circuit is simple, the output trigger pulse is safe and reliable, precision of real time control is high, stability and reliability that can bigger raising device; 3. has wide market application prospect.
(4) description of drawings
Accompanying drawing 1 for the related a kind of GAAA algorithm application based on genetic algorithm and ant group algorithm fusion of the present invention in the method for work overall flow synoptic diagram of the DSP of multi-storied garage parking position scheduling control system.
Accompanying drawing 2 is the GAAA algorithm flow chart of the genetic algorithm in the method for work of the related a kind of DSP control system that is applied to multi-storied garage parking position scheduling of the present invention with the ant group algorithm fusion.
(5) embodiment:
Embodiment: a kind of DSP control system that is applied to multi-storied garage parking position scheduling (is seen accompanying drawing 1,2) method of work, its working environment is by exchanging traversing motor at one, lifting motor, frequency converter, sensor, various mechanical switchs, wire rope, the parking stall framework, in the multi-storied garage that conventional equipment such as vehicle-containing and traversing guide rail constitutes, it is characterized in that: it is a senior decision making function of utilizing computing machine, genetic algorithm (Genetic Algorithm) and ant group algorithm (Ant Algorithm) combined constitute algorithm based on GAAA, be used for the method for multi-storied garage parking position scheduling by the DSP control system, it comprises following job step:
(1) confirm that the garage duty is ready: confirm earlier whether the garage is ready, can whether each equipment all can safe and reliable work in the garage, finish access car task with normal operating conditions;
(2) access car signal: receive access car signal in the garage, detect the garage internal memory state of picking up the car, when control system receives user's bicycle parking or during the signal of picking up the car, the DSP control system confirms that by the scanning to database this signal is bicycle parking or picks up the car;
(3) DSP control system start information is handled: the DSP control system begins to carry out information processing according to the GAAA algorithm of genetic algorithm and ant group algorithm fusion;
(4) parameter coding: at first, the direction of each parking stall action is carried out parameter coding, and utilize the rand function to generate the decimal system real coding population of some at random;
(5) setting of initial population: to colony's operation of evolving, setting the initial population data that some represent initial search point by genetic algorithm, is that the next stage output of optimal path code generates initial initial population;
(6) design of fitness function: promptly to the evaluation of hereditary ideal adaptation degree, be meant in genetic algorithm, determine that with the size of ideal adaptation degree this individuality is genetic to the probability in the colony of future generation; By the ability assessment to hereditary ideal adaptation environment, as the foundation of selection operation, it is to be formed by the objective function conversion; After fitness value function by design came out each chromosomal fitness value calculation, the individual inheritance of holding according to " survival of the fittest " principle was to the next generation, and bad individuality is eliminated; Then, OX crossover operator, the conversion mutation operator according to modified form in the genetic algorithm carries out calculated crosswise and conversion variation to it;
(7) setting of controlled variable: determine to the far-reaching parameter value of performance of genetic algorithms;
(8) asking for of GAAA algorithm and access car optimal route: the maximum-minimum ant MMAS of system algorithm is designed and improves, in the GAAA algorithm, ant algorithm adopts the maximum-minimum ant MMAS of system algorithm, and to some groups of optimization solutions through producing after the recursive iteration of setting algebraically in the step (5), adopt the maximum-minimum ant MMAS of system algorithm that these several groups of final outputs of optimization solution recursive iteration are preferably separated, promptly obtain the shortest path of vehicle access;
(9) conversion and control language and output control signal: convert the shortest path of the access vehicle obtained in the step (8) to control language, by DSP control system output signal, facilities such as control garage motor, frequency converter, vehicle-containing are finished the access car.
The direction of parking stall action is total in the above-mentioned said step (4): static, motionless, upwards, five kinds of states downwards, left and to the right; The coding method of adopting (seeing accompanying drawing 2) is: the odd bits of each individuality (chromosome) is represented the car item, and even bit is represented direction, respectively with 0,1,2,3,4 transfixions of representing the parking stall, make progress, this five kinds of states downwards, left and to the right; Form is as follows:
9 1 20 2 5 2 □□
The action of car item action car item action car item
Two kinds of methods are generally arranged: 1. according to the intrinsic knowledge of problem, manage to hold the distribution range of optimum solution take up space in whole problem space, set initial population in the setting (seeing accompanying drawing 2) of above-mentioned said step (5) initial population; 2. generate earlier the individuality of some at random, therefrom choose best individuality and be added in the initial population; The continuous iteration of this process is up to reaching predetermined scale; And through producing some groups of optimization solutions after the recursive iteration of setting algebraically.
The fitness of the individuality of the design (seeing accompanying drawing 2) of the middle fitness function of above-mentioned said step (6) is genetic to follow-on probability with this individuality and is directly proportional; The unique requirement of fitness function is that its result is a nonnegative value; The change of scale of fitness is certain mapping transformation to the objective function codomain, can overcome the prematurity convergence and roam phenomenon at random.
The individual inheritance of holding according to " survival of the fittest " principle in the above-mentioned said step (6) is divided into direct heredity or produces the heredity again of new individuality by intersecting to of future generation.
Design to the operation (seeing accompanying drawing 2) of genetic algorithm in the above-mentioned said step (6) comprises the design of selecting operator, crossover operator, mutation operator: 1. select operator: select operator also to be called and duplicate operator, be from contemporary community, to select some better individualities, and it is copied in the colony of future generation, the individual probability that duplicates is proportional to its adaptive value; 2. crossover operator: crossover operator is meant to be replaced the part-structure of two parent individualities reorganization and is generated new individual operation, the individuality that combination makes new advances, at strand space efficient search is arranged, reduce the failure probability to effective model simultaneously, various crossover operators all comprise two substances: determine the position of point of crossing and carry out the exchange of portion gene; 3. mutation operator: mutation operator is meant to be replaced some genic values in the individual coded strings with other genic value, form a new individuality, and the variation algorithm comprises the position of definitive variation point and carries out the genic value replacement.
The setting (seeing accompanying drawing 2) of controlled variable mainly comprises the long l of string, group size n, crossover probability p in the above-mentioned said step (7)
c, the variation Probability p
mDeng, wherein, the general parameter area of determining is: n=20～200, p
c=0.5～1.0, p
m=0～0.05;
Maximum in the above-mentioned said step (the 8)-minimum ant MMAS of system algorithm is a kind of based on bionic method, promptly calculate ability that ant relies on optimum path search and can find shortest path between ant nest and the food, comprise ant stay the disappearance of As time goes on can intensity volatilizing gradually of a kind of volatility secretion (pheromone) that is called pheromones and pheromones on the path of process; Ant can this material of perception in the process of looking for food existence and intensity thereof, and instruct the direction of motion of oneself with this, tend to move, promptly select the intensity of this material on the probability in this path and this paths at that time to be directly proportional towards the high direction of this material intensity; The path that pheromones intensity is high more selects its ant just many more, and then the intensity of the pheromones that stays on this path is just bigger, and the big pheromones of intensity attracts more ant, thereby form a kind of positive feedback,, make most ant all can walk this optimal path by this positive feedback.
Maximum-minimum ant the MMAS of system algorithm (seeing accompanying drawing 2) has three aspects different with the standard ant group algorithm in the above-mentioned said step (8):
1. pheromones update mode difference, each circulation back is only carried out pheromones to this circulation optimum solution or an ant up to the present finding out optimum solution and is upgraded, and in the standard ant group algorithm, all ants path of passing by is all carried out pheromones and upgraded, MMAS pheromones update mode is as follows:
τ
ij(t+n)＝ρτ
ij(t)+Δτ
ij best，Δτ
ij best＝1/f(s
best) (i)
F (s wherein
Best) expression iteration optimum solution (s
Ib) or globally optimal solution (s
Gb) value;
4. the stagnation for avoiding searching for, the codomain scope of the pheromones track amount on each element of separating is limited in [τ
Min, τ
Max] in the interval.That is: if τ is arranged
Ij(t)＞τ
Max, then put τ
Ij(t)=τ
MaxIf τ is arranged
Ij(t)＜τ
Min, then put τ
Ij(t)=τ
MinAnd pheromones track amount is not limited in standard ant system, make track amount on some paths far above other limits, thereby ant has stoped the more excellent behavior of separating of further search all along moving with paths;
5. for to make ant can search for new solution more, the pheromones Track Initiation is turned to τ in the starting stage of algorithm
Max, MMAS is owing to be limited in [τ to the codomain of the pheromones track amount on each path
Min, τ
Max] in the interval.
In the above-mentioned said step (8) maximum-minimum ant MMAS of system algorithm is designed and improve (seeing accompanying drawing 2), wherein the pheromones update mode to MMAS is: the first, and because of pheromones on each path among the MMAS is initialized as τ
Max, should reduce the pheromones on the relatively poor path rapidly, accelerate search speed, so after each circulation, calculate the mean value L of each paths length earlier
AveThen, increase less than the routing information of median length is plain, routing information element greater than median length reduces, second, the amount of the increase of each side information element or minimizing should be the same on each paths, and the length on each limit that should embody each paths is to the percentage contribution of path total length, so (i) formula is modified as following form:
τ
ij(t+n)＝ρτ
ij(t)+Δτ
ij，Δτ
ij＝∑Δτ
ij k (ii)
Δ τ wherein
Ij kRepresent that k ant stay quantity of information on the ij of path, Δ τ in this circulation
IjRepresent to stay in this circulation the informational capacity on the ij of path, wherein Δ τ
Ij kAs follows:
L wherein
kIt is the path that k ant passed by in this circulation; L
AveAverage length for m path that ant walks; d
IjBe the distance of city i to city j.
Claims (8)
1. method of work that is applied to the DSP control system of multi-storied garage parking position scheduling, its working environment is to exchange traversing motor at one by comprising, lifting motor, frequency converter, sensor, various mechanical switchs, wire rope, the parking stall framework, vehicle-containing and traversing guide rail are in the multi-storied garage that interior conventional equipment constitutes, it is characterized in that: it is a senior decision making function of utilizing computing machine, genetic algorithm (Genetic Algorithm) and ant group algorithm (Ant Algorithm) combined constitute algorithm based on GAAA, be used for the method for multi-storied garage parking position scheduling by the DSP control system, it comprises following job step:
(1) confirm that the garage duty is ready: confirm earlier whether the garage is ready, can whether each equipment all can safe and reliable work in the garage, finish access car task with normal operating conditions;
(2) access car signal: receive access car signal in the garage, detect the garage internal memory state of picking up the car, when control system receives user's bicycle parking or during the signal of picking up the car, the DSP control system confirms that by the scanning to database this signal is bicycle parking or picks up the car;
(3) DSP control system start information is handled: the DSP control system begins to carry out information processing according to the GAAA algorithm of genetic algorithm and ant group algorithm fusion;
(4) parameter coding: at first, the direction of each parking stall action is carried out parameter coding, and utilize the rand function to generate the decimal system real coding population of some at random;
(5) setting of initial population: to colony's operation of evolving, setting the initial population data that some represent initial search point by genetic algorithm, is that the next stage output of optimal path code generates initial initial population;
(6) design of fitness function: promptly to the evaluation of hereditary ideal adaptation degree, be meant in genetic algorithm, determine that with the size of ideal adaptation degree this individuality is genetic to the probability in the colony of future generation; By the ability assessment to hereditary ideal adaptation environment, as the foundation of selection operation, it is to be formed by the objective function conversion; After fitness value function by design came out each chromosomal fitness value calculation, the individual inheritance of holding according to " survival of the fittest " principle was to the next generation, and bad individuality is eliminated; Then, OX crossover operator, the conversion mutation operator according to modified form in the genetic algorithm carries out calculated crosswise and conversion variation to it;
(7) setting of controlled variable: determine to the far-reaching parameter value of performance of genetic algorithms;
(8) asking for of GAAA algorithm and access car optimal route: the maximum-minimum ant MMAS of system algorithm is designed and improves, in the GAAA algorithm, ant algorithm adopts the maximum-minimum ant MMAS of system algorithm, and utilize the initial population data that configure in the step (5) to carry out the recursive iteration of algebraically, to producing some groups of optimization solutions after the recursive iteration, adopt the maximum-minimum ant MMAS of system algorithm that these several groups of final outputs of optimization solution recursive iteration are preferably separated, promptly obtain the shortest path of vehicle access;
(9) conversion and control language and output control signal: convert the shortest path of the access vehicle obtained in the step (8) to control language, by comprising DSP control system output signal, control garage motor, frequency converter, vehicle-containing are finished the access car at interior facility.
2. according to the said a kind of method of work that is applied to the DSP control system of multi-storied garage parking position scheduling of claim 1, it is characterized in that the direction of parking stall action in the said step (4) is total: transfixion, upwards, five kinds of states downwards, left and to the right; The coding method of adopting is: each individuality is that said each chromosomal odd bits is represented the car item in claim 1 step (6), even bit is represented direction, respectively with 0,1,2,3,4 transfixions of representing the parking stall, upwards, this five kinds of states downwards, left and to the right.
3. according to the said a kind of method of work that is applied to the DSP control system of multi-storied garage parking position scheduling of claim 1, it is characterized in that in the setting of said step (5) initial population two kinds of methods being arranged generally: 1. according to the intrinsic knowledge of problem, manage to hold the distribution range of optimum solution take up space in whole problem space, set initial population; 2. generate earlier the individuality of some at random, therefrom choose best individuality and be added in the initial population; The continuous iteration of this process is up to reaching predetermined scale; And through producing some groups of optimization solutions after the recursive iteration of setting algebraically.
4. according to the said a kind of method of work that is applied to the DSP control system of multi-storied garage parking position scheduling of claim 1, it is characterized in that the fitness of the individuality of the design of fitness function is genetic to follow-on probability with this individuality in the said step (6) to be directly proportional; The unique requirement of fitness function is that its result is a nonnegative value.
5. according to the said a kind of method of work that is applied to the DSP control system of multi-storied garage parking position scheduling of claim 1, it is characterized in that the individual inheritance of holding according to " survival of the fittest " principle in the said step (6) is divided into direct heredity or produces the heredity again of new individuality by intersecting to of future generation.
6. according to the said a kind of method of work that is applied to the DSP control system of multi-storied garage parking position scheduling of claim 1, it is characterized in that in the said step (6) design of the fitness function of genetic algorithm is comprised the design of selecting operator, crossover operator, mutation operator: 1. select operator: select operator also to be called and duplicate operator, be from contemporary community, to select some better individualities, and it is copied in the colony of future generation, the individual probability that duplicates is proportional to its adaptive value; 2. crossover operator: crossover operator is meant to be replaced the part-structure of two parent individualities reorganization and is generated new individual operation, the individuality that combination makes new advances, at strand space efficient search is arranged, reduce the failure probability to effective model simultaneously, various crossover operators all comprise two substances: determine the position of point of crossing and carry out the exchange of portion gene; 3. mutation operator: mutation operator is meant to be replaced some genic values in the individual coded strings with other genic value, form a new individuality, and the variation algorithm comprises the position of definitive variation point and carries out the genic value replacement.
7. according to the said a kind of method of work that is applied to the DSP control system of multi-storied garage parking position scheduling of claim 1, it is characterized in that controlled variable mainly comprises the long l of string, group size n, crossover probability p in the said step (7)
c, the variation Probability p
m, wherein, the general parameter area of determining is: n=20～200, p
c=0.5～1.0, p
m=0～0.05, the long l that goes here and there sets according to actual conditions.
8. according to the said a kind of method of work that is applied to the DSP control system of multi-storied garage parking position scheduling of claim 1, it is characterized in that the maximum-minimum ant MMAS of the system algorithm in the said step (8) is a kind of based on bionic method, promptly calculate ability that ant relies on optimum path search and can find shortest path between ant nest and the food, comprise ant stay the disappearance of As time goes on can intensity volatilizing gradually of a kind of volatility secretion (pheromone) that is called pheromones and pheromones on the path of process; Ant can this material of perception in the process of looking for food existence and intensity thereof, and instruct the direction of motion of oneself with this, tend to move, promptly select the intensity of this material on the probability in this path and this paths at that time to be directly proportional towards the high direction of this material intensity; The path that pheromones intensity is high more selects its ant just many more, and then the intensity of the pheromones that stays on this path is just bigger, and the big pheromones of intensity attracts more ant, thereby form a kind of positive feedback,, make most ant all can walk this optimal path by this positive feedback.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN2008101528029A CN101504795B (en) | 2008-11-03 | 2008-11-03 | Working method for DSP control system applied to multi-storied garage parking position scheduling |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN2008101528029A CN101504795B (en) | 2008-11-03 | 2008-11-03 | Working method for DSP control system applied to multi-storied garage parking position scheduling |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN101504795A CN101504795A (en) | 2009-08-12 |
| CN101504795B true CN101504795B (en) | 2010-12-15 |
Family
ID=40977026
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN2008101528029A Expired - Fee Related CN101504795B (en) | 2008-11-03 | 2008-11-03 | Working method for DSP control system applied to multi-storied garage parking position scheduling |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN101504795B (en) |
Families Citing this family (17)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102044011B (en) * | 2009-10-20 | 2013-07-31 | 北京交通大学 | Method and system for scheduling police force resource |
| JP4959774B2 (en) * | 2009-11-30 | 2012-06-27 | インターナショナル・ビジネス・マシーンズ・コーポレーション | Application generation system, method and program |
| CN102663574B (en) * | 2012-03-23 | 2014-10-08 | 合肥工业大学 | Multi-RGV dynamic scheduling method based on genetic algorithm |
| CN102708698B (en) * | 2012-06-12 | 2014-07-02 | 北京理工大学 | Vehicle optimal-path navigation method based on vehicle internet |
| CN104765690B (en) * | 2015-04-22 | 2017-10-27 | 哈尔滨工业大学 | Embedded software test data generation method based on fuzzy genetic algorithm |
| CN105868854A (en) * | 2016-03-29 | 2016-08-17 | 贵州航天南海科技有限责任公司 | Optimal control method of parking lot path of stereoscopic parking device |
| CN106200577B (en) * | 2016-07-08 | 2018-10-30 | 青岛华通能源投资有限责任公司 | A kind of stereo garage dispatch control method |
| CN106651086B (en) * | 2016-09-07 | 2020-06-26 | 河南科技学院 | Automatic stereoscopic warehouse scheduling method considering assembly process |
| CN106682254B (en) * | 2016-09-30 | 2020-01-14 | 银江股份有限公司 | Stereo garage optimization method based on adaptive genetic algorithm |
| CN106651042B (en) * | 2016-12-27 | 2021-02-02 | 贵州航天南海科技有限责任公司 | Obstacle avoidance optimization path control method for parking spaces of three-dimensional parking equipment |
| CN106874551B (en) * | 2017-01-11 | 2020-12-01 | 成都信息工程大学 | Parallel parking method based on third-order arc tangent function model |
| CN107943045A (en) * | 2017-12-08 | 2018-04-20 | 江苏商贸职业学院 | A kind of method for planning path for mobile robot based on ant colony genetic fusion algorithm |
| CN108711300A (en) * | 2018-05-25 | 2018-10-26 | 深圳市易成自动驾驶技术有限公司 | Swarm intelligence is parked method, system and computer readable storage medium |
| CN110086855B (en) * | 2019-03-29 | 2021-11-05 | 江苏开拓信息与系统有限公司 | Intelligent Spark task perception scheduling method based on ant colony algorithm |
| CN110334838B (en) * | 2019-04-11 | 2023-06-23 | 国网新疆电力有限公司营销服务中心(资金集约中心、计量中心) | AGV trolley cooperative scheduling method and system based on ant colony algorithm and genetic algorithm |
| CN111898206B (en) * | 2020-07-30 | 2024-04-12 | 长安大学 | Parameter optimization method based on improved genetic algorithm, computer equipment and storage medium |
| CN113192215B (en) * | 2021-03-25 | 2022-09-13 | 湖南亿马智能科技有限公司 | Parking system based on JAVA platform |
-
2008
- 2008-11-03 CN CN2008101528029A patent/CN101504795B/en not_active Expired - Fee Related
Also Published As
| Publication number | Publication date |
|---|---|
| CN101504795A (en) | 2009-08-12 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN101504795B (en) | Working method for DSP control system applied to multi-storied garage parking position scheduling | |
| CN108764555B (en) | Shared bicycle parking point site selection method based on Hadoop | |
| CN106444755A (en) | Mobile robot path planning method and system based on improved genetic algorithm | |
| CN102854880B (en) | Robot whole-situation path planning method facing uncertain environment of mixed terrain and region | |
| CN109506655B (en) | Improved ant colony path planning algorithm based on non-uniform modeling | |
| CN110047278B (en) | Adaptive traffic signal control system and method based on deep reinforcement learning | |
| CN104156584A (en) | Sensor target assignment method and system for multi-objective optimization differential evolution algorithm | |
| CN107370188A (en) | A kind of power system Multiobjective Scheduling method of meter and wind power output | |
| CN105929848A (en) | Track planning method for multi-unmanned plane system in three-dimensional environment | |
| CN106200577A (en) | A kind of multi-storied garage dispatch control method | |
| CN112036540B (en) | Sensor number optimization method based on double-population hybrid artificial bee colony algorithm | |
| CN110162041A (en) | A kind of robot path planning method based on self-adapted genetic algorithm | |
| CN110345960B (en) | Route planning intelligent optimization method for avoiding traffic obstacles | |
| CN106193734A (en) | A kind of multi-storied garage dispatching control device | |
| CN104392283A (en) | Artificial fish swarm algorithm based traffic route searching method | |
| CN102262702B (en) | Decision-making method for maintaining middle and small span concrete bridges | |
| CN110232461A (en) | More interconnection vector machine water quality prediction methods based on quantum genetic algorithm optimization | |
| CN106600959A (en) | Traffic congestion index-based prediction method | |
| CN102843743A (en) | Hierarchical routing method of wireless sensor network based on flora foraging optimization algorithm | |
| CN101916312A (en) | Ship route intelligent generation method based on combination of artificial immunity and ant colony algorithm | |
| CN113917925B (en) | Mobile robot path planning method based on improved genetic algorithm | |
| CN104616084A (en) | Assembly sequence planning method | |
| CN115859785A (en) | Multi-objective integrated optimization method for goods allocation and job scheduling of stereoscopic warehouse | |
| Abramovich et al. | Multi-objective topology and weight evolution of neuro-controllers | |
| CN103592920A (en) | Hybrid flow shop scheduling method with finite buffers |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C14 | Grant of patent or utility model | ||
| GR01 | Patent grant | ||
| CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20101215 Termination date: 20141103 |
|
| EXPY | Termination of patent right or utility model |