CN108596359A - Rehabilitation doctor's advice task assigns optimization method, device and equipment - Google Patents
Rehabilitation doctor's advice task assigns optimization method, device and equipment Download PDFInfo
- Publication number
- CN108596359A CN108596359A CN201810217247.7A CN201810217247A CN108596359A CN 108596359 A CN108596359 A CN 108596359A CN 201810217247 A CN201810217247 A CN 201810217247A CN 108596359 A CN108596359 A CN 108596359A
- Authority
- CN
- China
- Prior art keywords
- particle
- value
- fitness
- local search
- doctor
- 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 34
- 238000005457 optimization Methods 0.000 title claims abstract description 33
- 239000002245 particle Substances 0.000 claims abstract description 323
- 238000013178 mathematical model Methods 0.000 claims abstract description 58
- 238000011156 evaluation Methods 0.000 claims abstract description 13
- 230000035772 mutation Effects 0.000 claims description 21
- 230000006870 function Effects 0.000 claims description 20
- 238000012545 processing Methods 0.000 claims description 18
- 230000000739 chaotic effect Effects 0.000 claims description 10
- 108090000623 proteins and genes Proteins 0.000 claims description 6
- 238000004590 computer program Methods 0.000 claims description 5
- 230000006978 adaptation Effects 0.000 claims description 3
- 238000003860 storage Methods 0.000 claims description 3
- 238000004422 calculation algorithm Methods 0.000 abstract description 63
- 238000009826 distribution Methods 0.000 abstract description 10
- 230000007246 mechanism Effects 0.000 description 14
- 230000002776 aggregation Effects 0.000 description 6
- 238000004220 aggregation Methods 0.000 description 6
- 238000004891 communication Methods 0.000 description 6
- 238000010586 diagram Methods 0.000 description 6
- 230000003044 adaptive effect Effects 0.000 description 5
- 230000036541 health Effects 0.000 description 4
- 238000012549 training Methods 0.000 description 4
- 238000004364 calculation method Methods 0.000 description 3
- 230000008859 change Effects 0.000 description 3
- 238000005516 engineering process Methods 0.000 description 3
- 235000013305 food Nutrition 0.000 description 3
- 238000007726 management method Methods 0.000 description 3
- 230000008569 process Effects 0.000 description 3
- 239000003814 drug Substances 0.000 description 2
- 230000002708 enhancing effect Effects 0.000 description 2
- 230000001965 increasing effect Effects 0.000 description 2
- 238000012804 iterative process Methods 0.000 description 2
- 238000002560 therapeutic procedure Methods 0.000 description 2
- 230000001133 acceleration Effects 0.000 description 1
- 238000013459 approach Methods 0.000 description 1
- 239000012141 concentrate Substances 0.000 description 1
- 238000010276 construction Methods 0.000 description 1
- 238000013499 data model Methods 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 230000006872 improvement Effects 0.000 description 1
- 238000007620 mathematical function Methods 0.000 description 1
- 235000012054 meals Nutrition 0.000 description 1
- 230000005012 migration Effects 0.000 description 1
- 238000013508 migration Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000003032 molecular docking Methods 0.000 description 1
- 230000005405 multipole Effects 0.000 description 1
- 238000002360 preparation method Methods 0.000 description 1
- 238000003825 pressing Methods 0.000 description 1
- 238000012913 prioritisation Methods 0.000 description 1
- 230000000750 progressive effect Effects 0.000 description 1
- 230000035484 reaction time Effects 0.000 description 1
- 238000011084 recovery Methods 0.000 description 1
- 238000012163 sequencing technique 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
- G06Q10/04—Forecasting or optimisation specially adapted for administrative or management purposes, e.g. linear programming or "cutting stock problem"
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/004—Artificial life, i.e. computing arrangements simulating life
- G06N3/006—Artificial life, i.e. computing arrangements simulating life based on simulated virtual individual or collective life forms, e.g. social simulations or particle swarm optimisation [PSO]
-
- G—PHYSICS
- G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
- G16H—HEALTHCARE INFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR THE HANDLING OR PROCESSING OF MEDICAL OR HEALTHCARE DATA
- G16H40/00—ICT specially adapted for the management or administration of healthcare resources or facilities; ICT specially adapted for the management or operation of medical equipment or devices
- G16H40/20—ICT specially adapted for the management or administration of healthcare resources or facilities; ICT specially adapted for the management or operation of medical equipment or devices for the management or administration of healthcare resources or facilities, e.g. managing hospital staff or surgery rooms
Landscapes
- Engineering & Computer Science (AREA)
- Business, Economics & Management (AREA)
- Health & Medical Sciences (AREA)
- General Business, Economics & Management (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Strategic Management (AREA)
- Biomedical Technology (AREA)
- Human Resources & Organizations (AREA)
- Economics (AREA)
- General Health & Medical Sciences (AREA)
- Software Systems (AREA)
- Artificial Intelligence (AREA)
- Computing Systems (AREA)
- Mathematical Physics (AREA)
- Molecular Biology (AREA)
- Evolutionary Computation (AREA)
- Data Mining & Analysis (AREA)
- Development Economics (AREA)
- Computational Linguistics (AREA)
- Game Theory and Decision Science (AREA)
- Biophysics (AREA)
- General Engineering & Computer Science (AREA)
- Entrepreneurship & Innovation (AREA)
- Marketing (AREA)
- Operations Research (AREA)
- Quality & Reliability (AREA)
- Tourism & Hospitality (AREA)
- Life Sciences & Earth Sciences (AREA)
- Epidemiology (AREA)
- Medical Informatics (AREA)
- Primary Health Care (AREA)
- Public Health (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention discloses a kind of rehabilitation doctor's advice tasks to assign optimization method, device and equipment, by the mathematical model for establishing doctor's advice task assignment problem, the mathematical model is converted into PSO Algorithm, the inertia weight and accelerator coefficient calculate particle is adjusted, and particle is handled into row variation by median particle, the individual extreme value of particle to meeting preset local search probability carries out local search, and the fitness of the particle of local search is evaluated, according to the presently described individual extreme value of evaluation result update and global extremum, determine rehabilitation doctor's advice task assignment, to realize the high efficiency and accuracy that improve the distribution of rehabilitation doctor's advice task, and achieve the purpose that patient receives rehabilitation in time.
Description
Technical field
The present invention relates to intelligent doctor's advice dispatch technique fields, assign optimization side more particularly to a kind of rehabilitation doctor's advice task
Method, device and electronic equipment.
Background technology
As the medical science of recovery therapy of one of four big pith of modern medicine system, just towards specialized, standardization and multipole
The trend development of change.The standardization of rehabilitation department therapeutic process is implemented accordingly to need more accurate management and control, and rehabilitation at present is cured
The informatization for the treatment of business falls behind relatively, and information system is only capable of realizing that rehabilitation doctor's advice set meal is opened in institute, simply charges and holds
Row result paper record, but rehabilitation is a very long process, needs rehabilitation doctor, treatment teacher, nurse and patient
Between exchange closely, need to record a large amount of medical data, wherein since patient needs under the attending physician transmitted according to nurse
The rehabilitation doctor's advice reached carries out rehabilitation, needs the accurate assignment for ensureing rehabilitation doctor's advice, and provide rehabilitation service to patient on time,
Collecting the feedback opinion of patient's treatment in time can just make patient achieve the purpose that preferable rehabilitation, therefore rehabilitation doctor's advice task
Assignment problem become current rehabilitation medical in a more important link.
The settling mode of the assignment problem of rehabilitation doctor's advice usually treatment teacher, which takes the lead, at present contacts rehabilitation doctor understanding doctor
The refinement progress advised, then treating physician's personnel's duty arrangement of different business group is seeked advice from, it is finally dispatched to specific people manually, and
The personnel are notified to provide rehabilitation service before the deadline;Or rehabilitation doctor is contacted using distribution on line i.e. treatment teacher and is existed
Enterprising practise medicine of system advises refinement and submission, and therapist actively gets task on this system;Or the assignment of doctor's advice task is asked
Then the shifts arrangement that topic is converted to therapist carries out doctor's advice task assignment.
But all there is certain defects for the method for distribution of existing rehabilitation doctor's advice.Participant's time is linked up under line,
Task assignment is carried out on line, which needs that staff's plenty of time is spent to link up rehabilitation doctor, therapist and caregiver
Member docking, it is difficult to meet medical advances pay close attention to and information sharing;Task is claimed on line, and rehabilitation doctor is opened in system
Vertical doctor's advice, and refined and submitted, treatment personnel claim accordingly in system according to the working time of oneself and service ability
Task, which will appear task and claims the case where delaying treatment not in time;Convert doctor's advice task assignment problem to row
Class's problem, which carry out when on-line optimization that system reaction time is slightly slow, and are not combined with manpower attendance situation, to needs
It the processing underactions of special circumstances such as works overtime, ask for leave and is late.Therefore, how efficiently and accurately rehabilitation doctor's advice task to be carried out
Assign to ensure that patient carries out the technical issues of rehabilitation is those skilled in the art's urgent need solution in time.
Invention content
It is directed to the above problem, a kind of rehabilitation doctor's advice task of present invention offer assigns optimization method, device and rehabilitation training
System realizes the high efficiency and accuracy for improving the distribution of rehabilitation doctor's advice task, and reaches patient and receive rehabilitation in time
Purpose.
To achieve the goals above, the present invention provides the following technical solutions:
A kind of rehabilitation doctor's advice task assignment optimization method, including:
Establish the mathematical model of doctor's advice task assignment problem;
The position and speed for initializing each particle in particle cluster, each particle is calculated according to the mathematical model
Fitness value determines individual extreme value and global extremum;
The inertia weight and accelerator coefficient of the particle cluster are adjusted according to the fitness of the particle, and according to
Preset median particle into row variation processing, recalculates the suitable of adjustment and variation treated each particle to particle
Angle value is answered, current individual extreme value and global extremum are preserved;
The individual extreme value of particle to meeting preset local search probability carries out local search, and to the grain of local search
The fitness of son is evaluated, according to the presently described individual extreme value of evaluation result update and global extremum;
Judge whether updated current global extremum meets preset end condition, if it is, according to the overall situation
Extreme value determines rehabilitation doctor's advice task assignment.
Preferably, the mathematical model for establishing doctor's advice task assignment problem, including:
Build the mathematical model of rehabilitation doctor's advice task assignment problem;
The constraints of the mathematical model is set, wherein the mathematical model is expressed asIt is described
Constraints includes
And dj∈ [0,24], cijThe cost spent, x are completed by j-th of people by i-th of taskijFor Boolean variable, m is medical institutions
Interior energy provides the personnel amount of rehabilitation service, tijI-th of task the time it takes cost, d are completed for j-th of peoplejIt is j-th
Working duration on the day of people.
Preferably, the position and speed of each particle in the initialization particle cluster, according to the mathematical model meter
The fitness value of each particle is calculated, determines individual extreme value and global extremum, including:
Determine the initial position and initial velocity of any particle;
When the particle position is migrated, according to formula X 'i=Xi+Vi' be updated to obtain by the initial position
Updated position, wherein the initial position is expressed as Xi, updated position is expressed as X 'i, Vi' it is any particle
Speed of the i in next iteration;
According to preset formula Vi'=ω Vi+c1rand(Yi-Xi)+c2rand(Yg-Xi), determine individual extreme value and global pole
Value, wherein ω is inertia weight, and rand is the random number in [0,1], YiFor individual extreme value, YgFor global extremum, c1 and c2 divide
Accelerator coefficient is not indicated;
The fitness value of each particle is determined according to the mathematical model.
Preferably, the fitness according to the particle carries out the inertia weight and accelerator coefficient of the particle cluster
Adjustment, and according to preset median particle to particle into row variation processing, recalculate adjustment and variation treated and is described
The fitness value of each particle preserves current individual extreme value and global extremum, including:
The fitness value of each particle is ranked up, median particle is determined according to ranking results;
According to the fitness value of the fitness value of each particle and the median particle, be calculated aggregation because
Son, wherein the clumping factor is expressed as gt,N indicates population scale, and t is iteration
Step number, k are Dynamic gene, fi tBe in group i-th of particle the t times iteration fitness value,It is the suitable of median particle
Answer angle value;
Tuning function is generated according to the clumping factor;
The inertia weight and accelerator coefficient are adjusted according to the Tuning function, the inertia weight after being adjusted
And accelerator coefficient;
Judge whether the fitness value of each particle is less than the fitness value of the median particle, if it is, to institute
It states particle and carries out Gaussian mutation, if it is not, then carrying out chaotic mutation to the particle;
Each particle is calculated according to the inertia weight and accelerator coefficient after adjustment, and according to variation treated particle
Fitness value, and preserve current individual extreme value and global extremum.
Preferably, the individual extreme value of the described pair of particle for meeting preset local search probability carries out local search, and right
The fitness of the particle of local search is evaluated, according to the presently described individual extreme value of evaluation result update and global extremum, packet
It includes:
Local search probability is calculated according to the individual extreme value and global extremum of current particle;
Judge whether the local search probability of each particle is more than predetermined threshold value, if it is, to particle carry out office
Portion is searched for;
The fitness of the particle after local search is updated, and updates the individual extreme value of the particle after the local search and complete
Office's extreme value.
A kind of rehabilitation doctor's advice task assignment optimization device, including:
Module is established, the mathematical model for establishing doctor's advice task assignment problem;
Initialization module, the position and speed for initializing each particle in particle cluster, according to the mathematical modulo
Type calculates the fitness value of each particle, determines individual extreme value and global extremum;
Adjust module, for according to the fitness of the particle to the inertia weight of the particle cluster and accelerator coefficient into
Row adjustment, and according to preset median particle to particle into row variation processing, recalculate adjustment and variation treated institute
The fitness value of each particle is stated, current individual extreme value and global extremum are preserved;
Local search module, the individual extreme value for the particle to meeting preset local search probability carry out part and search
Rope, and the fitness of the particle of local search is evaluated, according to the presently described individual extreme value of evaluation result update and the overall situation
Extreme value;
Determining module, for judging whether updated current global extremum meets preset end condition, if it is,
Rehabilitation doctor's advice task assignment is determined according to the global extremum.
A kind of computer readable storage medium is stored thereon with computer program, and the computer program is by processing equipment
Realize that the rehabilitation doctor's advice task described in preceding claim assigns optimization method when execution.
A kind of task assignment equipment, the equipment include:
Memory, for storing program;
Processor, for running described program, when the processor runs described program, the processor realizes
It states the rehabilitation doctor's advice task and assigns optimization method.
Compared to the prior art, since rehabilitation doctor's advice task assignment problem is a combinatorial optimization problem, alternative compared with
It is more, and constraints is complex.Rehabilitation is cured so proposing realized based on improved Particle Swarm Algorithm in the present invention
Advise task intelligence assign, by analyze rehabilitation doctor's advice task assignment problem the characteristics of, founding mathematical models, according to the mathematical modulo
Type feature selection particle cluster algorithm optimizes to obtain global optimum's extreme value, and according to the global optimum, extreme value determines rehabilitation doctor's advice
Task assignment.Since entire rehabilitation doctor's advice task assignment is determined and assigns on line by particle cluster algorithm
The time for communication that can save medical staff, and the foundation of mathematical model is to be based on various situations as constraints,
Consider emergency case so that processing is more flexible, ensure that the Accuracy and high efficiency that task is assigned, therefore can ensure to suffer from
Person carries out rehabilitation in time.Meanwhile inertia weight and accelerator coefficient adjustment have been carried out for particle cluster algorithm in the present invention,
And adaptive adjustment has been carried out to algorithm, it is various then so that covariation mechanism increases population by introducing median particle
Property, avoiding algorithm from being absorbed in, office is excellent, and local search mechanism accelerates the convergence rate in algorithm later stage, improves the search of particle cluster algorithm
Precision.
Description of the drawings
In order to more clearly explain the embodiment of the invention or the technical proposal in the existing technology, to embodiment or will show below
There is attached drawing needed in technology description to be briefly described, it should be apparent that, the accompanying drawings in the following description is only this
The embodiment of invention for those of ordinary skill in the art without creative efforts, can also basis
The attached drawing of offer obtains other attached drawings.
Fig. 1 is the flow diagram that a kind of rehabilitation doctor's advice task provided in an embodiment of the present invention assigns optimization method;
Fig. 2 is the flow of a kind of individual extreme value of determining particle cluster provided in an embodiment of the present invention and the side of global extremum
Schematic diagram;
Fig. 3 is a kind of flow diagram of local search approach provided in an embodiment of the present invention;
Fig. 4 is a kind of flow diagram improving particle cluster algorithm provided in an embodiment of the present invention;
Fig. 5 is the structural schematic diagram that a kind of rehabilitation doctor's advice task provided in an embodiment of the present invention assigns optimization device;
Fig. 6 is a kind of structural schematic diagram of task assignment equipment provided in an embodiment of the present invention.
Specific implementation mode
Following will be combined with the drawings in the embodiments of the present invention, and technical solution in the embodiment of the present invention carries out clear, complete
Site preparation describes, it is clear that described embodiments are only a part of the embodiments of the present invention, instead of all the embodiments.It is based on
Embodiment in the present invention, it is obtained by those of ordinary skill in the art without making creative efforts every other
Embodiment shall fall within the protection scope of the present invention.
Term " first " and " second " in description and claims of this specification and above-mentioned attached drawing etc. are to be used for area
Not different objects, rather than for describing specific sequence.In addition term " comprising " and " having " and their any deformations,
It is intended to cover and non-exclusive includes.Such as it contains the process of series of steps or unit, method, system, product or sets
It is standby not to be set in the step of having listed or unit, but the step of may include not listing or unit.
An embodiment of the present invention provides a kind of doctor's advice tasks to assign optimization method, and referring to attached drawing 1, this method includes:
S11, the mathematical model for establishing doctor's advice task assignment problem;
In the present embodiment, task assignment problem of the doctor's advice task assignment problem primarily directed to rehabilitation doctor's advice.Due to health
Multiple doctor's advice task assignment problem is related to multiple links and multiple operational staffs, therefore rehabilitation doctor's advice task assignment problem can return
It receives as combinatorial optimization problem, is commonly described as:There is n task to need to assign m people to complete, each task can only be dispatched to one
Individual, and everyone can only execute a task in the same time, all tasks are assigned the personnel amount for completing to specify and cannot be surpassed
It crosses in the rehabilitation medical mechanism and the personnel amount of rehabilitation service is provided, and be dispatched to the sum of someone's task execution time in one day
No more than the working duration of the personnel, and certain rehabilitation doctor's advice tasks can only can be completed by the talent with certain skills.
Since the cost that position level and the time cost similarities and differences cause different people to execute different task costs is also different.Rehabilitation is cured
It is exactly to find a kind of assignment to advise the optimization aim of task assignment problem so that the Least-cost of cost.
Wherein, in order to reach so that the Least-cost that rehabilitation doctor's advice assignment is spent, in the present embodiment, can provide one
The method that kind establishes the mathematical model of the problem, specifically, S11 includes:
By involved by rehabilitation doctor's advice task assignment problem variable and each constraints be converted into cost function, build health
The mathematical model of multiple doctor's advice task assignment problem, the mathematical model can be expressed as the form of formula (1):
The constraints of the mathematical model (1) is set, wherein constraints include following formula (2), (3), (4),
(5)、(6):
dj∈[0,24](6)
cijThe cost spent, x are completed by j-th of people by i-th of taskijFor Boolean variable, and meet when i points of task
X when being completed to personnel jij=1, other situations xij=0;M provides the personnel amount of rehabilitation service, t for medical institutions' interior energyijFor
J-th of people completes i-th of task the time it takes cost, djWorking duration on the day of for j-th of people, theoretic boundary model
It is 0 to 24 hours to enclose, and ideal value is 8 hours, supports the special circumstances such as work overtime and ask for leave.
During actual optimization, for cijCalculating can introduce the reward factor and penalty factor, as the people of certain task assignment
When member is the personnel that patient specially requires, target function value can be rewarded;When the task execution time for being dispatched to someone
The sum of with the same day working duration deviate it is larger when, target function value can be punished, purpose is exactly to make final assignment
Execution can improve patient satisfaction and the work incentive of employee.
After the mathematical model for creating rehabilitation doctor's advice task assignment problem, since the mathematical model is combinatorial optimization problem
Mathematical function, be in embodiments of the present invention by particle cluster algorithm optimization seek optimal solution.
The position and speed of S12, each particle in initialization particle cluster, each grain is calculated according to the mathematical model
The fitness value of son determines individual extreme value and global extremum;
A kind of method of the individual extreme value and global extremum of determining particle cluster is additionally provided in embodiments of the present invention, is joined
See that Fig. 2, S12 include:
S121, the initial position and initial velocity for determining any particle;
S122, when the particle position is migrated, update particle position, i.e.,:
According to formula X 'i=Xi+Vi' be updated the initial position to obtain updated position, wherein it is described first
Beginning position is expressed as Xi, updated position is expressed as X 'i, Vi' it is speed of any particle i in next iteration;
S123, according to preset formula Vi'=ω Vi+c1rand(Yi-Xi)+c2rand(Yg-Xi), determine individual extreme value and complete
Office's extreme value, wherein ω is inertia weight, and rand is the random number in [0,1], YiFor individual extreme value, YgFor global extremum, c1 and
C2 indicates accelerator coefficient respectively;
S124, the fitness value that each particle is determined according to the mathematical model.
The mechanism of particle cluster algorithm optimization is that population scale represents the number of particle, and each particle represents a potential solution,
Food is exactly global optimum position, and population constantly migrates to food, until finding food.If search space is n dimensions, first
Determine that the initial position and initial velocity of i-th of particle, initial position can be expressed as X in S121i=(xi1,xi2,...,
xin), initial velocity can be expressed as Vi=(vi1,vi2,...,vin);
In S122, particle position constantly migrates, and is updated to the position and speed of particle, and the wherein update of position is public
Formula (7) is as follows:
X′i=Xi+Vi' (7)
Wherein Vi' be i-th of particle next iteration speed, formula (8) is as follows:
Vi'=ω Vi+c1rand(Yi-Xi)+c2rand(Yg-Xi) (8)
Wherein, ω is inertia weight, determines particle last iteration speed to current influence degree;Rand is [0,1]
Interior random number, YiFor individual extreme value, YgFor global extremum, c1And c2Accelerator coefficient, c are indicated respectively1Indicate particle by individual
Influence degree, c2Indicate influence degree of the particle by group.Therefore can according to the initial position and speed of particle, and
Position and speed after migration extrapolates the individual extreme value and global extremum of the particle cluster, and determines the fitness of particle.
S13, the inertia weight and accelerator coefficient of the particle cluster are adjusted according to the fitness of the particle, and
Adjustment and variation treated each particle are recalculated into row variation processing to particle according to preset median particle
Fitness value, preserve current individual extreme value and global extremum;
S13 is specifically included in an embodiment of the present invention:
The fitness value of each particle is ranked up, median particle is determined according to ranking results;
According to the fitness value of the fitness value of each particle and the median particle, be calculated aggregation because
Son, wherein the clumping factor is expressed as gt,N indicates population scale, and t is iteration
Step number, k are Dynamic gene, fi tBe in group i-th of particle the t times iteration fitness value,It is the suitable of median particle
Answer angle value;
Tuning function is generated according to the clumping factor;
The inertia weight and accelerator coefficient are adjusted according to the Tuning function, the inertia weight after being adjusted
And accelerator coefficient;
Judge whether the fitness value of each particle is less than the fitness value of the median particle, if it is, to institute
It states particle and carries out Gaussian mutation, if it is not, then carrying out chaotic mutation to the particle;
Each particle is calculated according to the inertia weight and accelerator coefficient after adjustment, and according to variation treated particle
Fitness value, and preserve current individual extreme value and global extremum.
Specifically, the principle of modified particle swarm optiziation proposed by the present invention is to make full use of particle group in iteration mistake
Distribution situation in journey proposes that clumping factor is used for the aggregation extent of quantitative description particle.It is adaptively adjusted using clumping factor
The inertia weight and accelerator coefficient of whole algorithm improve the search precision of algorithm, and combine median particle, and increasing, population is various
Property while, utilize the local search ability of Gaussian mutation, accelerate searching algorithm efficiency;If being inferior to median particle, to the grain
Son carries out chaotic mutation, using the ergodic of chaos, increases population diversity, avoids algorithm from being absorbed in office excellent.
The clumping factor proposed in the embodiment of the present invention is expressed as gt, formula (9) is as follows:
Wherein, N indicates population scale, and t is iterative steps, and k is Dynamic gene, fi tBe in group i-th of particle at the t times
The fitness value of iteration,It is the fitness value of median particle.
Median particle is to carry out the fitness value of particle in population after sorting from small to large, will be in ranking results
The particle that fitness value is located at the centre position in entire sequence is determined as median particle.
Difference is larger from formula (9) as can be seen that between algorithm iteration initial stage, particle, and clumping factor levels off to 0;
In the iteration later stage, particle is gradually close to global optimum's particle, and the fitness value difference between particle is smaller, and clumping factor becomes at this time
It is bordering on 1.
In particle cluster algorithm, the adjustment of inertia weight and accelerator coefficient has decisive influence, existing skill to algorithm performance
It is by inertia weight in art according to iterative steps Serial regulation, accelerator coefficient uses empirical value, and this adjustment mode is all adopted for 2
With unified pattern without representativeness, in embodiments of the present invention according to the distribution situation in population iterative process, dynamically certainly
Adjustment algorithm parameter is adapted to, Tuning function is proposed according to clumping factor, it is as follows:
Wherein,Indicate the adaptive optimal control angle value of the population in the t times iteration, fi tIt is i-th of particle in the t times iteration
In ideal adaptation angle value.
Inertia weight ω is the important parameter of particle cluster algorithm, its value of iteration initial stage is answered larger, and the overall situation for enhancing algorithm is opened
Hair ability, the iteration later stage, its value should be smaller, and the local exploring ability for enhancing algorithm obtains inertia weight according to Tuning function
More new formula it is as follows:
Accelerator coefficient c1It is close to individual extreme value direction to influence particle, accelerator coefficient c2Particle is influenced to global optimum position
It is close, so at algorithm initial stage, c1Should be larger, c2Should be smaller, population diversity, iteration later stage c can be increased in this way1It answers
It is smaller, c2It answers larger, algorithmic statement can be accelerated, accelerator coefficient is updated according to Tuning function, specifically:
The c in formula (12)1maxAnd c1minAccelerator coefficient c is indicated respectively1Maximum value and minimum value;In formula (13)
c2maxAnd c2minAccelerator coefficient c is indicated respectively2Maximum value and minimum value.
Particle cluster algorithm in an iterative process, can be quickly close to global optimum's particle and lack population diversity by particle,
Algorithm is set to be absorbed in office excellent.And be typically that solve annual reporting law, to be absorbed in office excellent by the average value of particle fitness in the prior art,
But the average value of particle fitness can not react the current optimizing situation of population well.It proposes in embodiments of the present invention
Median particle, that is, the fitness value size for pressing particle sort, and take out the fitness value of intermediate particulate to react entire population
Distribution situation, judged for the fitness value of the fitness value and median particle of each particle, that is, use formula
(14):
Wherein, fi tIndicate fitness value of i-th of particle in the t times iteration,Indicate median particle at the t times repeatedly
For when fitness value.
Judged for each particle according to formula (14), if meeting formula (14) is better than median particle, is made
With Gaussian mutation, using the local search ability of Gaussian mutation, variation has generated the particle of the position of attachment, accelerates algorithm and explores
Ability;If being unsatisfactory for formula (14), that is, it is inferior to median particle, then uses chaotic mutation, the characteristics of using chaos ergodic, with
Machine variation generates the particle of position farther out, increases population diversity.
Chaotic mutation and the formula of Gaussian mutation difference are as follows:
In formula (15)The value for indicating t step iterative chaotic variations, works as u=4,And
When, chaos phenomenon will be generated,The traversal in (0,1).
δ is the standard deviation of Gaussian distributions in formula (16), and μ is it is expected, the smaller distributions of δ can more concentrate on x=u
Position, on the contrary can more it disperse.
It is as follows to the formula of particle position into row variation:
For particle into after row variation, the position before variation and variable position line emphasis are exactly the new position after variation, formula
It is as follows:
In embodiments of the present invention, it is exactly covariation to the essence of particle into row variation, i.e., by each particle and middle position
Number particles be compared, if be better than median particle, illustrate the particle closer to global optimum position, it should around it into
Row is explored;If being inferior to median particle, illustrate that the particle is farther from global optimum position, random site change should be carried out to it
It is different, bigger solution space is explored, two kinds of Variation mechanism collaborations carry out, switch in real time, when the particle for being listed in median carries out chaos change
After different, fitness has been better than median particle, then both ensure that algorithm in this way with regard to carrying out Gaussian mutation in make a variation next time
Diversity, and accelerate the exploration efficiency of algorithm.
S14, particle to meeting preset local search probability individual extreme value carry out local search, and to local search
The fitness of particle evaluated, according to the presently described individual extreme value of evaluation result update and global extremum;
Referring to Fig. 3, step S14 includes:
S141, local search probability is calculated according to the individual extreme value and global extremum of current particle;
S142, judge whether the local search probability of each particle is more than predetermined threshold value, if it is, executing S143;
S143, local search is carried out to the particle;
The fitness of particle after S144, update local search, and update the individual pole of the particle after the local search
Value and global extremum.
Since particle cluster algorithm is in later stage of evolution, convergence rate is slack-off, to carry out local search, and local search is searched with the overall situation
Suo Butong, local search population size is smaller, and the iterative steps of local search are fewer than global search very much, specially will be complete
Office's search is divided into early period, mid-term and later stage by iterative steps, the iterative steps of local search global search early period, mid-term and after
Phase, respectively 2%, 6%, the 10% of global search iterative steps.
If carrying out local search to each particle, can also it take a significant amount of time, and to the particle of fitness value difference
Local search is carried out, algorithmic statement is unfavorable for, reduces algorithm search efficiency, so when carrying out local search, for each
A local search probability L is arranged in particleim, formula is as follows:
Wherein, gtFor clumping factor, k is constant, depending on value is by specific mathematical model and population scale size, at this
Preferably 0.9 in inventive embodiments, then local search probability LimIt is the number between [0.5,1].
For each particle, proposes the Rule of judgment of a local search, could be provided as:
Lis=min (Lim,m)
Wherein, m is constant, and value is an empirical value, in the present invention depending on Population Size and the mathematical model of optimization
Embodiment in can choose 0.8.Work as LisWhen equal to 0.8, local search is carried out to the individual extreme value of the particle, at this time entire group
Body aggregation extent is big, and the particle is close to global optimum position;Conversely, entire group collection of drama degree is little, the particle is far from global
Optimal location, corresponding local search probability is smaller, can't carry out local search to the particle.
Then, after local search more new particle fitness, and the individual extreme value and global extremum of more new particle.Particle
Fitness be value of the particle in object function, object function can be the function determined in mathematical model.
S15, judge whether updated current global extremum meets preset end condition, if it is, executing S16;
S16, rehabilitation doctor's advice task assignment is determined according to the global extremum.
Since rehabilitation doctor's advice task assignment problem is a combinatorial optimization problem, alternative is more, and constraints
It is complex.So proposing the intelligence point realized based on improved Particle Swarm Algorithm to rehabilitation doctor's advice task in the present invention
Group, by analyze rehabilitation doctor's advice task assignment problem the characteristics of, founding mathematical models, according to the mathematical model feature select particle
Group's algorithm optimizes to obtain global optimum's extreme value, and according to the global optimum, extreme value determines rehabilitation doctor's advice task assignment.
Due to entire rehabilitation doctor's advice task assignment be determined and assign on line by particle cluster algorithm can save doctor
The time for communication of shield personnel, and the foundation of mathematical model is to be based on various situations as constraints, it is contemplated that burst feelings
Condition so that processing is more flexible, ensure that the Accuracy and high efficiency that task is assigned, therefore can ensure that patient carries out health in time
Multiple treatment.Meanwhile inertia weight and accelerator coefficient adjustment have been carried out for particle cluster algorithm in the present invention, and algorithm is carried out
Then adaptive adjustment makes the covariation mechanism to increase population diversity, avoids algorithm by introducing median particle
It is absorbed in that office is excellent, local search mechanism accelerates the convergence rate in algorithm later stage, improves the search precision of particle cluster algorithm.
Referring to Fig. 4, a kind of improvement particle cluster algorithm is additionally provided in the embodiment of the present invention, which includes:
S401, setting algorithm parameter, wherein algorithm parameter includes:Particle number N, inertia weight ωmax、ωmin, accelerate
Index boundaries c1max、c1min、c2max、c2min, greatest iteration step number Tmax, constant k, m;
S402, particle populations are initialized, the speed V of random initializtion particle, position X, and limits speed and position
The value range set;
S403, fitness evaluation is carried out to particle, calculates the fitness value of each particle, and determined and obtain global optimum position
It sets;
S404, clumping factor, median particle fitness value and local searching probability is calculated, and updates inertia power
Speed and the position of weight, accelerator coefficient and particle cluster algorithm;
S405, it is directed to each particle, judges whether to meet Gaussian mutation condition, if it is, executing S406, otherwise executed
S407;
S406, Gaussian mutation is carried out to particle;
S407, chaotic mutation is carried out to particle;
S408, fitness evaluation, more new individual extreme value and global extremum are carried out to the particle after variation;
S409, judge whether to meet local search condition, if satisfied, then executing S410;
S410, local search is carried out to the particle;
S411, the fitness of the particle of progress local search is evaluated, more new individual extreme value and global extremum;
S412, judge whether algorithm meets end condition, if satisfied, S413 is executed, if not satisfied, then executing S404;
S413, output global extremum are as optimal case.
It is easily trapped into that local better solution, local search ability be poor, the algorithm later stage for particle cluster algorithm in the present embodiment
The slack-off problem of convergence rate, the improved APSO algorithm of proposition, covariation extreme value increase population diversity,
Avoiding algorithm from being absorbed in, office is excellent, and local search mechanism enhances the local search ability of algorithm, improves search precision;According to population
The self-adaptive step random search method mechanism that the aggregation extent of distribution proposes, accelerates the convergence rate in algorithm search later stage.
In actual application, for rehabilitation training system, it will usually be divided into doctor's advice collection subsystem, doctor's advice assigns son
System, doctor's advice executive subsystem and statistics sub system.
Wherein, doctor's advice collection subsystem, the rehabilitation doctor's advice for being opened to patient according to doctor and doctor's advice type masterplate, it is raw
At standardized doctor's advice information;
Doctor's advice assigns subsystem, for generating rehabilitation doctor's advice task assignment, doctor's advice according to standardized doctor's advice information
The rehabilitation doctor's advice task provided during subsystem is executed to the embodiment of the present invention is provided and assigns optimization method;
Doctor's advice executive subsystem, for being distributed rehabilitation doctor's advice in preset time according to rehabilitation doctor's advice task assignment
To operational staff corresponding with rehabilitation doctor's advice, and record execution information of the operational staff to rehabilitation doctor's advice;
Statistics sub system for counting operational staff to the execution information of rehabilitation doctor's advice, and generates statistical report form,
Statistical report form is sent to specified administrative staff.
For example, physiatrician can open rehabilitation doctor's advice in rehabilitation training system, it is assumed that within certain following week, health
Multiple doctor is that 10 patients have opened 15 rehabilitation doctor's advices, this 15 rehabilitation doctor's advices can be dispatched to 5 rehabilitation therapists and come
At.
The information such as doctor's advice information, including the treatment project of doctor's advice, treatment time are transferred to doctor's advice to assign in subsystem, together
When by following one week of therapist arrange an order according to class and grade information and professional service direction is also communicated to doctor's advice and assigns in subsystem, system will doctor
It advises the demand of execution and service that therapist provides, is expressed with data model and constraints, improved population is utilized to calculate
Method optimizes assignment, at this point, number of particles N is 515, it is therapist that 15 doctor's advices, which are randomly assigned to 5, as the first of particle
Beginning position x, the maximum value that inertia weight is arranged are 0.9, minimum value 0.5, and the maximum value of two accelerator coefficients is 2.1, minimum value
It is 1.8, iterative steps 100.
After each iteration optimization, next iteration doctor's advice task dispatching priority is evaluated according to work saturation degree, is worked
Saturation degree is a constraints of the mathematical model, and on the basis of meeting other constraintss, rehabilitation doctor's advice is preferentially assigned
Give work saturation degree low treating physician, wherein work saturation degree can be evaluated by the saturation degree formula that works, specifically
's:
Work saturation degree=employee arranges an order according to class and grade working hour/and (employee arranges an order according to class and grade working hour-employee work hour is provided)
Then optimal case is found according to the particle cluster algorithm of optimization, assignment is automatically generated, if certain treating physician
It asks for leave temporarily, can adjust the information of arranging an order according to class and grade that doctor's advice is assigned, then regenerate prioritization scheme, assignment adjustment can also be carried out manually,
The rehabilitation doctor's advice task of the employee is directly adjusted to an other employee.
After assignment finally determines, doctor's advice executive subsystem can be entered, the subsystem prompt same day needs to execute
Rehabilitation task, while the notification mechanisms for thering is task delay to execute.After the completion of work, by statistics sub system, management level can be clear
The process of work for grasping all employees, greatly promotes the efficiency of management.
The rehabilitation training system provided through the embodiment of the present invention, online doctor's advice task in real time are assigned, and medical matters people is relieved
The heavy communication work that member's manual tasks are assigned improves the working efficiency of medical staff, and system is assigned in time automatically, is avoided
Because confusing communication cause task execution not in time the problem of;System can select the personnel of current most competitiveness when assigning automatically,
Raising personnel's utilization rate, while appropriate personnel can also improve the satisfaction of patient to provide suitably service;Task is evaded
Assign and arrange an order according to class and grade conflict the problem of, to working overtime, asking for leave and the processing of interim emergency situations is more flexible, guarantee has most suitable always
Personnel to patient provide rehabilitation service.
A kind of rehabilitation doctor's advice task assignment optimization device is additionally provided in embodiments of the present invention, referring to Fig. 5, including:
Module 501 is established, the mathematical model for establishing doctor's advice task assignment problem;
Initialization module 502, the position and speed for initializing each particle in particle cluster, according to the mathematics
Model calculates the fitness value of each particle, determines individual extreme value and global extremum;
Module 503 is adjusted, for being to the inertia weight of the particle cluster and acceleration according to the fitness of the particle
Number is adjusted, and is handled into row variation particle according to preset median particle, after recalculating adjustment and variation processing
Each particle fitness value, preserve current individual extreme value and global extremum;
Local search module 504, the individual extreme value for the particle to meeting preset local search probability carry out part
Search, and the fitness of the particle of local search is evaluated, according to the presently described individual extreme value of evaluation result update and entirely
Office's extreme value;
Determining module 505, for judging whether updated current global extremum meets preset end condition, if
It is that rehabilitation doctor's advice task assignment is then determined according to the global extremum.
Optionally, the module of establishing includes:
Construction unit, the mathematical model for building rehabilitation doctor's advice task assignment problem;
Constraints setting unit, the constraints for the mathematical model to be arranged, wherein the mathematical model is expressed asThe constraints includes
And dj∈ [0,24], cijThe cost spent, x are completed by j-th of people by i-th of taskijFor Boolean variable, m is medical institutions
Interior energy provides the personnel amount of rehabilitation service, tijI-th of task the time it takes cost, d are completed for j-th of peoplejIt is j-th
Working duration on the day of people.
Optionally, the initialization module includes:
Originally determined unit, initial position and initial velocity for determining any particle;
Location updating unit, for when the particle position is migrated, according to formula X 'i=Xi+Vi' will be described first
Beginning position is updated to obtain updated position, wherein the initial position is expressed as Xi, updated position is expressed as X
′i, Vi' it is speed of any particle i in next iteration;
Extreme value determination unit, for according to preset formula Vi'=ω Vi+c1rand(Yi-Xi)+c2rand(Yg-Xi), it determines
Individual extreme value and global extremum, wherein ω is inertia weight, and rand is the random number in [0,1], YiFor individual extreme value, YgIt is complete
Office's extreme value, c1 and c2 indicate accelerator coefficient respectively;
Fitness determination unit, the fitness value for determining each particle according to the mathematical model.
Optionally, the adjustment module includes:
Sequencing unit is ranked up for the fitness value to each particle, and median is determined according to ranking results
Particle;
Clumping factor computing unit is used for the adaptation of the fitness value and the median particle according to each particle
Clumping factor is calculated in angle value, wherein the clumping factor is expressed as gt,N tables
Show that population scale, t are iterative steps, k is Dynamic gene, fi tBe in group i-th of particle the t times iteration fitness value,It is the fitness value of median particle;
Function generation unit, for generating Tuning function according to the clumping factor;
Adjustment unit is adjusted for being adjusted to the inertia weight and accelerator coefficient according to the Tuning function
Inertia weight after whole and accelerator coefficient;
Judging unit, for judging whether the fitness value of each particle is less than the fitness value of the median particle,
If it is, Gaussian mutation is carried out to the particle, if it is not, then carrying out chaotic mutation to the particle;
Numerical calculation unit, for according to the inertia weight and accelerator coefficient after adjustment, and according to variation treated grain
Son calculates the fitness value of each particle, and preserves current individual extreme value and global extremum.
Optionally, the local search module includes:
Probability calculation unit, it is general for local search to be calculated according to the individual extreme value and global extremum of current particle
Rate;
Judging unit is searched for, for judging whether the local search probability of each particle is more than predetermined threshold value, if it is,
Local search is carried out to the particle;
Extreme value updating unit, the fitness for updating the particle after local search, and after updating the local search
The individual extreme value and global extremum of particle.
Device through this embodiment, it is proposed that the intelligence to rehabilitation doctor's advice task is realized based on improved Particle Swarm Algorithm
Assign, by analyze rehabilitation doctor's advice task assignment problem the characteristics of, founding mathematical models, according to the mathematical model feature select grain
Swarm optimization optimizes to obtain global optimum's extreme value, and according to the global optimum, extreme value determines rehabilitation doctor's advice task assignment side
Case.Since entire rehabilitation doctor's advice task assignment is capable of saving of being determined and assign on line by particle cluster algorithm
The time for communication of medical staff, and the foundation of mathematical model is to be based on various situations as constraints, it is contemplated that burst
Situation so that processing is more flexible, ensure that the Accuracy and high efficiency that task is assigned, therefore can ensure that patient carries out in time
Rehabilitation.Meanwhile carried out inertia weight and accelerator coefficient for particle cluster algorithm in the present invention and adjusted, and to algorithm into
It has gone adaptive adjustment, has then made covariation mechanism increase population diversity by introducing median particle, avoid calculating
Method is absorbed in that office is excellent, and local search mechanism accelerates the convergence rate in algorithm later stage, improves the search precision of particle cluster algorithm.
An embodiment of the present invention provides a kind of computer readable storage mediums, are stored thereon with computer program, the meter
Calculation machine program realizes that above-mentioned rehabilitation doctor's advice task assigns optimization method when being executed by processing equipment.
An embodiment of the present invention provides a kind of task assignment equipments, and referring to Fig. 6, which includes:
Memory 601, for storing program;
Processor 602, for running described program, specifically, processor realizes following steps when executing program:
Establish the mathematical model of doctor's advice task assignment problem;
The position and speed for initializing each particle in particle cluster, each particle is calculated according to the mathematical model
Fitness value determines individual extreme value and global extremum;
The inertia weight and accelerator coefficient of the particle cluster are adjusted according to the fitness of the particle, and according to
Preset median particle into row variation processing, recalculates the suitable of adjustment and variation treated each particle to particle
Angle value is answered, current individual extreme value and global extremum are preserved;
The individual extreme value of particle to meeting preset local search probability carries out local search, and to the grain of local search
The fitness of son is evaluated, according to the presently described individual extreme value of evaluation result update and global extremum;
Judge whether updated current global extremum meets preset end condition, if it is, according to the overall situation
Extreme value determines rehabilitation doctor's advice task assignment.
Optionally, the mathematical model for establishing doctor's advice task assignment problem, including:
Build the mathematical model of rehabilitation doctor's advice task assignment problem;
The constraints of the mathematical model is set, wherein the mathematical model is expressed asIt is described
Constraints includes With
dj∈ [0,24], cijThe cost spent, x are completed by j-th of people by i-th of taskijFor Boolean variable, m is in medical institutions
The personnel amount of rehabilitation service, t can be providedijI-th of task the time it takes cost, d are completed for j-th of peoplejFor j-th of people
The working duration on the same day.
Optionally, the position and speed of each particle in the initialization particle cluster, according to the mathematical model meter
The fitness value of each particle is calculated, determines individual extreme value and global extremum, including:
Determine the initial position and initial velocity of any particle;
When the particle position is migrated, according to formula X 'i=Xi+Vi' be updated to obtain by the initial position
Updated position, wherein the initial position is expressed as Xi, updated position is expressed as X 'i, Vi' it is any particle
Speed of the i in next iteration;
According to preset formula Vi'=ω Vi+c1rand(Yi-Xi)+c2rand(Yg-Xi), determine individual extreme value and global pole
Value, wherein ω is inertia weight, and rand is the random number in [0,1], YiFor individual extreme value, YgFor global extremum, c1 and c2 divide
Accelerator coefficient is not indicated;
The fitness value of each particle is determined according to the mathematical model.
Optionally, the fitness according to the particle carries out the inertia weight and accelerator coefficient of the particle cluster
Adjustment, and according to preset median particle to particle into row variation processing, recalculate adjustment and variation treated and is described
The fitness value of each particle preserves current individual extreme value and global extremum, including:
The fitness value of each particle is ranked up, median particle is determined according to ranking results;
According to the fitness value of the fitness value of each particle and the median particle, be calculated aggregation because
Son, wherein the clumping factor is expressed as gt,N indicates population scale, and t is iteration
Step number, k are Dynamic gene, fi tBe in group i-th of particle the t times iteration fitness value,It is the suitable of median particle
Answer angle value;
Tuning function is generated according to the clumping factor;
The inertia weight and accelerator coefficient are adjusted according to the Tuning function, the inertia weight after being adjusted
And accelerator coefficient;
Judge whether the fitness value of each particle is less than the fitness value of the median particle, if it is, to institute
It states particle and carries out Gaussian mutation, if it is not, then carrying out chaotic mutation to the particle;
Each particle is calculated according to the inertia weight and accelerator coefficient after adjustment, and according to variation treated particle
Fitness value, and preserve current individual extreme value and global extremum.
Optionally, the individual extreme value of the described pair of particle for meeting preset local search probability carries out local search, and right
The fitness of the particle of local search is evaluated, according to the presently described individual extreme value of evaluation result update and global extremum, packet
It includes:
Local search probability is calculated according to the individual extreme value and global extremum of current particle;
Judge whether the local search probability of each particle is more than predetermined threshold value, if it is, to particle carry out office
Portion is searched for;
The fitness of the particle after local search is updated, and updates the individual extreme value of the particle after the local search and complete
Office's extreme value.
Equipment through this embodiment, it is proposed that the intelligence to rehabilitation doctor's advice task is realized based on improved Particle Swarm Algorithm
Assign, by analyze rehabilitation doctor's advice task assignment problem the characteristics of, founding mathematical models, according to the mathematical model feature select grain
Swarm optimization optimizes to obtain global optimum's extreme value, and according to the global optimum, extreme value determines rehabilitation doctor's advice task assignment side
Case.Since entire rehabilitation doctor's advice task assignment is capable of saving of being determined and assign on line by particle cluster algorithm
The time for communication of medical staff, and the foundation of mathematical model is to be based on various situations as constraints, it is contemplated that burst
Situation so that the more flexible Accuracy and high efficiency that ensure that task and assign of processing, therefore can ensure that patient carries out in time
Rehabilitation.Meanwhile carried out inertia weight and accelerator coefficient for particle cluster algorithm in the present invention and adjusted, and to algorithm into
It has gone adaptive adjustment, has then made covariation mechanism increase population diversity by introducing median particle, avoid calculating
Method is absorbed in that office is excellent, and local search mechanism accelerates the convergence rate in algorithm later stage, improves the search precision of particle cluster algorithm.
Each embodiment is described by the way of progressive in this specification, the highlights of each of the examples are with other
The difference of embodiment, just to refer each other for identical similar portion between each embodiment.For device disclosed in embodiment
For, since it is corresponded to the methods disclosed in the examples, so description is fairly simple, related place is said referring to method part
It is bright.
The foregoing description of the disclosed embodiments enables those skilled in the art to implement or use the present invention.
Various modifications to these embodiments will be apparent to those skilled in the art, as defined herein
General Principle can be realized in other embodiments without departing from the spirit or scope of the present invention.Therefore, of the invention
It is not intended to be limited to the embodiments shown herein, and is to fit to and the principles and novel features disclosed herein phase one
The widest range caused.
Claims (8)
1. a kind of rehabilitation doctor's advice task assigns optimization method, which is characterized in that including:
Establish the mathematical model of doctor's advice task assignment problem;
The position and speed for initializing each particle in particle cluster, the adaptation of each particle is calculated according to the mathematical model
Angle value determines individual extreme value and global extremum;
The inertia weight and accelerator coefficient of the particle cluster are adjusted according to the fitness of the particle, and according to default
Median particle to particle into row variation processing, recalculate the fitness of adjustment and variation treated each particle
Value preserves current individual extreme value and global extremum;
The individual extreme value of particle to meeting preset local search probability carries out local search, and to the particle of local search
Fitness is evaluated, according to the presently described individual extreme value of evaluation result update and global extremum;
Judge whether updated current global extremum meets preset end condition, if it is, according to the global extremum
Determine rehabilitation doctor's advice task assignment.
2. according to the method described in claim 1, it is characterized in that, the mathematical model for establishing doctor's advice task assignment problem,
Including:
Build the mathematical model of rehabilitation doctor's advice task assignment problem;
The constraints of the mathematical model is set, wherein the mathematical model is expressed asThe constraint
Condition includes
And dj∈ [0,24], cijThe cost spent, x are completed by j-th of people by i-th of taskijFor Boolean variable, m is medical institutions
Interior energy provides the personnel amount of rehabilitation service, tijI-th of task the time it takes cost, d are completed for j-th of peoplejIt is j-th
Working duration on the day of people.
3. according to the method described in claim 2, it is characterized in that, the position of each particle in the initialization particle cluster
And speed, the fitness value of each particle is calculated according to the mathematical model, determines individual extreme value and global extremum, including:
Determine the initial position and initial velocity of any particle;
When the particle position is migrated, according to formula Xi'=Xi+Vi' the initial position is updated is updated
Position afterwards, wherein the initial position is expressed as Xi, updated position is expressed as Xi', Vi' it is that any particle i exists
The speed of next iteration;
According to preset formula Vi'=ω Vi+c1rand(Yi-Xi)+c2rand(Yg-Xi), determine individual extreme value and global extremum,
In, ω is inertia weight, and rand is the random number in [0,1], YiFor individual extreme value, YgFor global extremum, c1 and c2 are indicated respectively
Accelerator coefficient;
The fitness value of each particle is determined according to the mathematical model.
4. according to the method described in claim 3, it is characterized in that, the fitness according to the particle is to the particle collection
The inertia weight and accelerator coefficient of group is adjusted, and is handled into row variation particle according to preset median particle, again
The fitness value for calculating adjustment and variation treated each particle preserves current individual extreme value and global extremum, including:
The fitness value of each particle is ranked up, median particle is determined according to ranking results;
According to the fitness value of the fitness value of each particle and the median particle, clumping factor is calculated,
In, the clumping factor is expressed as gt,N indicates population scale, and t is iterative steps, k
For Dynamic gene, fi tBe in group i-th of particle the t times iteration fitness value,It is the fitness of median particle
Value;
Tuning function is generated according to the clumping factor;
The inertia weight and accelerator coefficient are adjusted according to the Tuning function, the inertia weight after being adjusted and plus
Fast coefficient;
Judge whether the fitness value of each particle is less than the fitness value of the median particle, if it is, to the grain
Son carries out Gaussian mutation, if it is not, then carrying out chaotic mutation to the particle;
According to the inertia weight and accelerator coefficient after adjustment, and according to the suitable of variation treated particle calculates each particle
Angle value is answered, and preserves current individual extreme value and global extremum.
5. according to the method described in claim 4, it is characterized in that, described pair of particle for meeting preset local search probability
Individual extreme value carries out local search, and evaluates the fitness of the particle of local search, is updated according to evaluation result current
The individual extreme value and global extremum, including:
Local search probability is calculated according to the individual extreme value and global extremum of current particle;
Judge whether the local search probability of each particle is more than predetermined threshold value, is searched if it is, carrying out part to the particle
Rope;
The fitness of the particle after local search is updated, and updates the individual extreme value of the particle after the local search and global pole
Value.
6. a kind of rehabilitation doctor's advice task assigns optimization device, which is characterized in that including:
Module is established, the mathematical model for establishing doctor's advice task assignment problem;
Initialization module, the position and speed for initializing each particle in particle cluster, according to the mathematical model meter
The fitness value of each particle is calculated, determines individual extreme value and global extremum;
Module is adjusted, for being adjusted to the inertia weight and accelerator coefficient of the particle cluster according to the fitness of the particle
It is whole, and according to preset median particle to particle into row variation processing, recalculate adjustment and variation treated and is described each
The fitness value of a particle preserves current individual extreme value and global extremum;
Local search module, the individual extreme value for the particle to meeting preset local search probability carry out local search, and
The fitness of the particle of local search is evaluated, according to the presently described individual extreme value of evaluation result update and global extremum;
Determining module, for judging whether updated current global extremum meets preset end condition, if it is, according to
The global extremum determines rehabilitation doctor's advice task assignment.
7. a kind of computer readable storage medium, is stored thereon with computer program, which is characterized in that the computer program quilt
Realize that the rehabilitation doctor's advice task described in claim 1-5 assigns optimization method when processing equipment executes.
8. a kind of task assignment equipment, which is characterized in that the equipment includes:
Memory, for storing program;
Processor, for running described program, when the processor runs described program, the processor realizes right and wants
The rehabilitation doctor's advice task described in 1-5 is asked to assign optimization method.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201810217247.7A CN108596359A (en) | 2018-03-16 | 2018-03-16 | Rehabilitation doctor's advice task assigns optimization method, device and equipment |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201810217247.7A CN108596359A (en) | 2018-03-16 | 2018-03-16 | Rehabilitation doctor's advice task assigns optimization method, device and equipment |
Publications (1)
Publication Number | Publication Date |
---|---|
CN108596359A true CN108596359A (en) | 2018-09-28 |
Family
ID=63626500
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201810217247.7A Pending CN108596359A (en) | 2018-03-16 | 2018-03-16 | Rehabilitation doctor's advice task assigns optimization method, device and equipment |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN108596359A (en) |
Cited By (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110827946A (en) * | 2019-10-31 | 2020-02-21 | 泰康保险集团股份有限公司 | Rehabilitation advice refining method and system, electronic device and storage medium |
CN111048189A (en) * | 2019-11-20 | 2020-04-21 | 泰康保险集团股份有限公司 | Information recording method and apparatus, electronic device, and storage medium |
CN112863656A (en) * | 2021-01-15 | 2021-05-28 | 上海助院科技有限公司 | Automatic recording device and method for medical advice execution |
Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN102222268A (en) * | 2011-06-02 | 2011-10-19 | 西安电子科技大学 | Method for scheduling flow shop based on multi-swarm hybrid particle swarm algorithm |
US20160292013A1 (en) * | 2012-12-10 | 2016-10-06 | Zte Corporation | Method and System for Scheduling Task in Cloud Computing |
CN107065876A (en) * | 2017-04-28 | 2017-08-18 | 西北工业大学 | Method for planning path for mobile robot based on Modified particle swarm optimization |
CN107547457A (en) * | 2017-09-15 | 2018-01-05 | 重庆大学 | A kind of approach for blind channel equalization based on Modified particle swarm optimization BP neural network |
-
2018
- 2018-03-16 CN CN201810217247.7A patent/CN108596359A/en active Pending
Patent Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN102222268A (en) * | 2011-06-02 | 2011-10-19 | 西安电子科技大学 | Method for scheduling flow shop based on multi-swarm hybrid particle swarm algorithm |
US20160292013A1 (en) * | 2012-12-10 | 2016-10-06 | Zte Corporation | Method and System for Scheduling Task in Cloud Computing |
CN107065876A (en) * | 2017-04-28 | 2017-08-18 | 西北工业大学 | Method for planning path for mobile robot based on Modified particle swarm optimization |
CN107547457A (en) * | 2017-09-15 | 2018-01-05 | 重庆大学 | A kind of approach for blind channel equalization based on Modified particle swarm optimization BP neural network |
Non-Patent Citations (1)
Title |
---|
苗长新等: "基于优化神经网络和DGA的变压器故障诊断", 《高压电器》 * |
Cited By (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110827946A (en) * | 2019-10-31 | 2020-02-21 | 泰康保险集团股份有限公司 | Rehabilitation advice refining method and system, electronic device and storage medium |
CN110827946B (en) * | 2019-10-31 | 2023-04-07 | 泰康保险集团股份有限公司 | Rehabilitation advice refining method and system, electronic device and storage medium |
CN111048189A (en) * | 2019-11-20 | 2020-04-21 | 泰康保险集团股份有限公司 | Information recording method and apparatus, electronic device, and storage medium |
CN112863656A (en) * | 2021-01-15 | 2021-05-28 | 上海助院科技有限公司 | Automatic recording device and method for medical advice execution |
CN112863656B (en) * | 2021-01-15 | 2022-12-02 | 上海助院科技有限公司 | Automatic recording device and method for medical order execution |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
Chaudhuri et al. | Fuzzy genetic heuristic for university course timetable problem | |
CN106055379A (en) | Method and system for scheduling computational task | |
CN108122179A (en) | Delamination Teaching cource arrangement method and device, delamination Teaching curricula-variable method and system | |
CN108596359A (en) | Rehabilitation doctor's advice task assigns optimization method, device and equipment | |
Lin | Context-aware task allocation for distributed agile team | |
Dhodiya et al. | Genetic algorithm based hybrid approach to solve fuzzy multi-objective assignment problem using exponential membership function | |
CN109214448A (en) | Non- good performance staff training method, system, terminal and computer readable storage medium | |
El-Qulity et al. | A generalized national planning approach for admission capacity in higher education: a nonlinear integer goal programming model with a novel differential evolution algorithm | |
CN114546608A (en) | Task scheduling method based on edge calculation | |
Whig et al. | Reinforcement Learning for Automated Medical Diagnosis and Dynamic Clinical Regimes | |
Zhu et al. | Scalable approximate policies for Markov decision process models of hospital elective admissions | |
Perzina et al. | Self-learning genetic algorithm for a timetabling problem with fuzzy constraints | |
Al-Yakoob et al. | A column generation mathematical programming approach for a class-faculty assignment problem with preferences | |
Lai et al. | An artificial intelligence approach to course timetabling | |
Yu et al. | Scheduling Multiobjective Dynamic Surgery Problems via $ Q $-Learning-Based Meta-Heuristics | |
WO2022214977A1 (en) | Assigning surgical operations to operating rooms | |
CN114781806A (en) | Multi-combination course arrangement method and device based on multi-objective evolutionary algorithm and readable medium | |
Guo et al. | MTIRL: Multi-trainer interactive reinforcement learning system | |
Gonsalves et al. | Artificial Immune Algorithm for exams timetable | |
Zhai et al. | Research on Application of Meticulous Nursing Scheduling Management Based on Data‐Driven Intelligent Optimization Technology | |
Saemi et al. | Solving task scheduling problem in mobile cloud computing using the hybrid multi-objective Harris Hawks optimization algorithm | |
CN115330277B (en) | Method and device for automatically selecting elevator by robot | |
Rurifandho et al. | Doctors dynamic scheduling for outpatient, inpatient, and surgery using genetic algorithm | |
Li et al. | A complex contract negotiation model based on hybrid intelligent algorithm | |
Ariyazand et al. | Optimization of a multi-objective university course timetabling problem with a hybrid WOA&NSGA-II (Case study: IAU, Robat Karim branch) |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
PB01 | Publication | ||
PB01 | Publication | ||
SE01 | Entry into force of request for substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
RJ01 | Rejection of invention patent application after publication |
Application publication date: 20180928 |
|
RJ01 | Rejection of invention patent application after publication |