WO2014173271A1 - 数字化人机界面监视单元数量优化方法及系统 - Google Patents

数字化人机界面监视单元数量优化方法及系统 Download PDF

Info

Publication number
WO2014173271A1
WO2014173271A1 PCT/CN2014/075824 CN2014075824W WO2014173271A1 WO 2014173271 A1 WO2014173271 A1 WO 2014173271A1 CN 2014075824 W CN2014075824 W CN 2014075824W WO 2014173271 A1 WO2014173271 A1 WO 2014173271A1
Authority
WO
WIPO (PCT)
Prior art keywords
fuzzy
monitoring
human
optimization factor
error
Prior art date
Application number
PCT/CN2014/075824
Other languages
English (en)
French (fr)
Inventor
张力
蒋建军
戴立操
李鹏程
胡鸿
黄卫刚
邹衍华
方小勇
戴忠华
王春辉
苏德颂
李晓蔚
Original Assignee
湖南工学院
南华大学
中广核核电运营有限公司
大亚湾核电运营管理有限责任公司
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by 湖南工学院, 南华大学, 中广核核电运营有限公司, 大亚湾核电运营管理有限责任公司 filed Critical 湖南工学院
Publication of WO2014173271A1 publication Critical patent/WO2014173271A1/zh

Links

Classifications

    • GPHYSICS
    • G05CONTROLLING; REGULATING
    • G05BCONTROL OR REGULATING SYSTEMS IN GENERAL; FUNCTIONAL ELEMENTS OF SUCH SYSTEMS; MONITORING OR TESTING ARRANGEMENTS FOR SUCH SYSTEMS OR ELEMENTS
    • G05B13/00Adaptive control systems, i.e. systems automatically adjusting themselves to have a performance which is optimum according to some preassigned criterion
    • G05B13/02Adaptive control systems, i.e. systems automatically adjusting themselves to have a performance which is optimum according to some preassigned criterion electric
    • G05B13/0265Adaptive control systems, i.e. systems automatically adjusting themselves to have a performance which is optimum according to some preassigned criterion electric the criterion being a learning criterion
    • G05B13/0275Adaptive control systems, i.e. systems automatically adjusting themselves to have a performance which is optimum according to some preassigned criterion electric the criterion being a learning criterion using fuzzy logic only
    • GPHYSICS
    • G05CONTROLLING; REGULATING
    • G05BCONTROL OR REGULATING SYSTEMS IN GENERAL; FUNCTIONAL ELEMENTS OF SUCH SYSTEMS; MONITORING OR TESTING ARRANGEMENTS FOR SUCH SYSTEMS OR ELEMENTS
    • G05B13/00Adaptive control systems, i.e. systems automatically adjusting themselves to have a performance which is optimum according to some preassigned criterion
    • G05B13/02Adaptive control systems, i.e. systems automatically adjusting themselves to have a performance which is optimum according to some preassigned criterion electric
    • G05B13/04Adaptive control systems, i.e. systems automatically adjusting themselves to have a performance which is optimum according to some preassigned criterion electric involving the use of models or simulators
    • GPHYSICS
    • G21NUCLEAR PHYSICS; NUCLEAR ENGINEERING
    • G21DNUCLEAR POWER PLANT
    • G21D3/00Control of nuclear power plant
    • G21D3/008Man-machine interface, e.g. control room layout
    • GPHYSICS
    • G21NUCLEAR PHYSICS; NUCLEAR ENGINEERING
    • G21DNUCLEAR POWER PLANT
    • G21D3/00Control of nuclear power plant
    • G21D3/04Safety arrangements
    • YGENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
    • Y02TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
    • Y02EREDUCTION OF GREENHOUSE GAS [GHG] EMISSIONS, RELATED TO ENERGY GENERATION, TRANSMISSION OR DISTRIBUTION
    • Y02E30/00Energy generation of nuclear origin

Definitions

  • the present invention relates to the field of digital control of nuclear power plants, and in particular to a method and system for optimizing the number of digital human-machine interface monitoring units applied to a digital control system of a nuclear power plant.
  • BACKGROUND OF THE INVENTION An operator obtains information on the current operating status of a factory through a human-machine interface, and performs the corresponding behavioral actions after the judgment and decision are made on the obtained information.
  • the design of human-machine interface will have an impact on a series of behaviors such as information acquisition and judgment.
  • accidents caused by human factors have taken the lead, and human accidents have become the main source of accidents such as nuclear power plants and chemical plants.
  • the existing method for optimizing the number of human-machine interface monitoring units mainly includes:
  • Finite differential method This method is the most primitive gradient method and requires multiple iterations of simulation, suitable for transient and regenerative simulation.
  • the finite-differentiation method requires more operations to obtain a more differential estimate in order to obtain a more reliable estimate, which results in a much higher operating cost.
  • the likelihood ratio method is also called the scoring function.
  • the estimation of the gradient can be obtained only by a single simulation run. The basic idea is to analyze the dependence of the probability measure of the system sample path on the random variable distribution function.
  • the likelihood ratio is obtained by the measure transformation, and the estimator of the performance measure is constructed. This method is also suitable for numerical problems; the method is obtained by some estimation.
  • the particle swarm optimization method uses the PSO method to solve the optimization problem.
  • the potential solution of each problem is defined as a particle 01 in the search space.
  • Each particle can be represented by a triple (xi, vi, pi ), where Xi represents the current position of the particle; vi represents the current velocity of the particle; pi represents the best position (individual experience) that the particle itself has searched for.
  • Unconstrained indirect method is to use the functional state to optimize by differential or variational, mainly including gradient method, Newton method and variable scale method.
  • the disadvantage of the gradient method is that the objective function must have a first-order partial derivative and needs to be calculated. The iterative point is far from the most advantageous, and the function value decreases rapidly. The closer to the most advantageous, the slower the convergence speed.
  • the disadvantage of Newton's method is that the objective function must have one-order, second-order partial derivatives and Hessian matrix non-singular and positive or negative. It is necessary to calculate the first-order, second-order partial derivatives and the inverse of the Hessian matrix. The program is complex and computationally intensive. .
  • variable scale method The disadvantage of the variable scale method is that the first-order partial derivative needs to be calculated, and the numerical stability is still not ideal due to the rounding error and the inaccuracy of the one-dimensional search. Sometimes the computational failure is caused by the singularity of the variable-scale matrix caused by the calculation error.
  • Constrained direct method This method is applicable to optimization problems involving only inequality constraints, such as network method, random direction search method and complex shape method.
  • Neural network method In fact, the neural network is rarely used in optimization.
  • the neural network is mainly used to correct the input weighting factor and process calculation.
  • This study is mainly used to optimize the number of monitoring units for digital human-machine interface, and does not involve the correction of weight factors.
  • Optimization The existing human-machine interface monitoring unit quantity optimization method has a large amount of data calculation and many iterations, and cannot meet the problem of rapidly designing the number of digital human-machine interface monitoring units, so that the operator can quickly and accurately obtain monitoring information and reduce the cause. The probability of human accidents caused by the acquisition of human-machine interface information.
  • the present invention aims to provide a digital human-machine interface monitoring unit number optimization method to solve the technical problem of quickly optimizing the number of digital human-machine interface monitoring units.
  • Another object of the present invention is to provide a digital human-machine interface monitoring unit number optimization system to solve the technical problem of quickly optimizing the number of digital human-machine interface monitoring units.
  • a digital human-machine interface monitoring unit quantity optimization method comprising the following steps: Fuzzyly segmenting the number of monitoring units of the digitized human-machine interface to be processed to constitute multiple fuzzy a fuzzy optimization factor sequence of the optimization factor; performing segmentation immune evolution on the fuzzy optimization factor sequence to extract a fuzzy optimization factor from the fuzzy optimization factor sequence; and sequentially monitoring the extracted fuzzy optimization factor to monitor a human error probability Calculating to obtain the extracted human error probability corresponding to the extracted fuzzy optimization factor; determining whether the monitoring person's error probability tends to be stable, that is, determining a plurality of extracted fuzzy optimization factors for monitoring the human error probability Whether the difference between the maximum value and the minimum value is less than a predetermined threshold, and if so, terminating the segmentation immune evolution of the fuzzy optimization factor sequence, if otherwise returning to the step of segmentation immune evolution of the fuzzy optimization factor sequence; Multiple monitors are compared for their probability of error, The fuzzy optimization factor with the smallest probability of error is
  • the step of calculating the error probability of the monitoring person includes: performing an affinity error rate calculation on the extracted fuzzy optimization factor by using a mapping function to calculate an affinity affecting the monitoring result of the transfer of Mrs. Marco's reliability model operation result
  • the error rate is obtained by monitoring the extracted fuzzy optimization factor by the reliability model operation to calculate the probability of monitoring the human error caused by the fuzzy optimization factor.
  • the mapping function used in the affinity error rate calculation is:
  • a digital human-machine interface monitoring unit quantity optimization system includes: a fuzzy segmentation module, wherein a number of monitoring units for a digitized human-machine interface to be processed is subjected to fuzzy segmentation to constitute a plurality of blurs a fuzzy optimization factor sequence of the optimization factor; a segmentation extraction module, configured to perform segmentation immune evolution on the fuzzy optimization factor sequence to extract a fuzzy optimization factor from the fuzzy optimization factor sequence; and a fault probability calculation module, which sequentially extracts The fuzzy optimization factor is used to monitor the human error probability calculation to obtain the extracted human error probability corresponding to the extracted fuzzy optimization factor; and the evolution determination module is configured to determine whether the monitoring human error probability tends to be stable, That is, determining whether the difference between the maximum value and the minimum value of the error probability of the plurality of
  • the error probability calculation module includes: an affinity error rate calculation unit, configured to perform an affinity error rate calculation on the extracted fuzzy optimization factor by using a mapping function, to calculate a reliability monitoring model for affecting the monitoring and transfer of Markov
  • the error probability calculation unit of the operation result is configured to perform the monitoring and transfer of the extracted fuzzy optimization factor by the reliability model calculation to calculate the probability of monitoring the human error caused by the fuzzy optimization factor.
  • the digital human-machine interface monitoring unit quantity optimization method of the invention establishes a fuzzy optimization factor sequence for monitoring the quantity of the monitoring unit by using fuzzy segmentation, and further adopts the segmentation immune evolution method in the fuzzy optimization factor
  • the fuzzy optimization factor is selected in the sequence to monitor the probability of human error, and the evolution process is terminated when the probability of human error is stabilized.
  • the number of iterations of evolution is reduced, the amount of computational data is reduced, and the computational efficiency is improved. Optimize the optimal range of monitoring units quickly and efficiently to ensure the operator's efficiency and quality of monitoring data, and improve the safety and reliability of digital control systems.
  • the digital human-machine interface monitoring unit quantity optimization system of the present invention establishes a fuzzy optimization factor sequence for monitoring the quantity of the monitoring unit by using fuzzy segmentation, and further selects the fuzzy optimization factor in the fuzzy optimization factor sequence by the segmentation immune evolution method.
  • the evolution process is terminated when the probability of error is monitored, the number of iterations of evolution is reduced, the amount of computational data is reduced, and the computational efficiency is improved, so that the optimal optimization can be optimized quickly and efficiently.
  • the range of the number of monitoring units is to ensure that the operator obtains the efficiency and quality of the monitoring data and improves the safety and reliability of the digital control system.
  • FIG. 1 is a flow chart showing the steps of a digital human-machine interface monitoring unit quantity optimization method according to a preferred embodiment of the present invention
  • FIG. 2 is a flow chart of step S30 of FIG. 1
  • FIG. 3 is a digital embodiment of the preferred embodiment of the present invention.
  • FIG. 4 is a schematic block diagram of the error probability calculation module of FIG.
  • the invention is used for solving the problem of the number fitness of the digital human-machine interface monitoring unit, so that the operator can quickly and accurately acquire the monitoring information through the human-machine interface, so as to improve the safety and reliability of the digital monitoring system. Referring to FIG.
  • a preferred embodiment of the present invention provides a method for optimizing the number of digitized human-machine interface monitoring units, which specifically includes the following steps: Step S10: Perform fuzzy segmentation on the number of monitoring units of the digitized human-machine interface to be processed and establish a fuzzy optimization factor sequence of multiple fuzzy optimization factors;
  • the fuzzy theory is mainly used to study and clarify some ambiguous problems in the real world.
  • the invention utilizes the fuzzy theory and combines and studies the object, that is, the number of monitoring units of the digital human-machine interface, and blurs the number of monitoring units to form a fuzzy optimization factor. Therefore, the fuzzy optimization factor encodes the number of segments of the number of monitoring units, and does not encode a specific number of values.
  • Step S20 Perform segmental immune evolution on the fuzzy optimization factor sequence to extract the fuzzy optimization factor from the fuzzy optimization factor sequence; the traditional optimization method can achieve higher speed and precision when solving small and medium-sized optimization problems, but in real life In most cases, the problem is NP-complete. Due to the complexity of the solution, the traditional optimization method can not adapt, because the "dimensional disaster" problem is inevitable.
  • the fuzzy optimization factor sequence based on the fuzzy segmentation is used to search for the fuzzy optimization factor by using the segmentation immune evolution method, and the fuzzy optimization factor obtained by the search is used to monitor the human error probability calculation.
  • the specific segmentation immune evolution method can be decomposed into the following steps: First, the fuzzy segments in the fuzzy optimization factor sequence are arranged in a small to large order, and the queues Q1 and Q2 are respectively initialized; The middle segment d_ml enters the queue Q1, and dequeues the queue Q1. At this time, the middle segment d_ml divides the original fuzzy optimization factor sequence into two sub-sequences; the fuzzy segment dequeued from the queue Q1 is divided into steps n and segments m.
  • the segment enters the queue Q2 in turn, and the fuzzy segment in the queue Q2 is dequeued, and the fuzzy segment is monitored in turn to monitor the human error probability until the queue Q2 is empty; the step of repeating the return to the construction queue Q1 is specifically After extracting the middle segment d_ml, the remaining fuzzy segment is divided into two sub-sequences by the middle segment d_ml, and the middle segments of the above two sub-sequences are respectively taken into the queue Q1; this is repeated until the calculated fuzzy segment monitors the error probability For stability, the difference between the maximum value and the minimum value of the monitored human error probability of the extracted fuzzy optimization factor is less than a preset threshold.
  • Step S30 sequentially monitor the extracted fuzzy optimization factor to monitor the human error probability calculation, to obtain the extracted fuzzy optimization factor to monitor the human error probability; and use the segmentation immune evolution method to extract the fuzzy in the fuzzy optimization factor sequence
  • the segment is the fuzzy optimization factor.
  • the steps for monitoring the fuzzy optimization factor to monitor the human error probability include: Step S31: Using the mapping function to perform the affinity error rate calculation on the extracted fuzzy optimization factor to calculate the impact monitoring Transfer the affinity error rate of Mrs. Marco's operation result due to the reliability model; where the mapping function of the affinity error rate operation is:
  • piMS ⁇ H DJ indicates the probability of error in monitoring the i-th target under the influence of human factors and decision-making behavior
  • piMS H DJMSi—J indicates the failure rate from the i-1th target to the i-th target under the influence of human influence factors and decision-making behaviors
  • p ⁇ MSilH im , Di ⁇ indicates the influence factors and decision-making behaviors in humans
  • p ⁇ H im ⁇ indicates the monitoring error rate caused by the human influence factor during monitoring
  • ⁇ ⁇ indicates the monitoring error rate caused by the decision behavior during monitoring.
  • the calculation of im ⁇ takes into account the fuzzy quantity optimization factor, the human influence factor and the monitoring time.
  • the calculation formula of p ⁇ H im ⁇ includes the affinity error rate. ⁇ ⁇ , so in the application monitoring transfer Mrs. Marco can calculate the probability of human error due to the reliability model calculation first need to calculate the affinity error rate Pij.
  • Step S40 determining whether the probability of the error of the monitoring person tends to be stable, that is, determining whether the difference between the maximum value and the minimum value of the error probability of the monitored fuzzy optimization factor is less than a predetermined threshold, and if so, terminating the fuzzy optimization factor sequence Segment immunological evolution, step S50 is performed; if not, then return to step S20 to continue segmentation immune evolution of the fuzzy optimization factor sequence to extract the fuzzy segment of the deep evolution as the fuzzy optimization factor.
  • Step S50 Comparing the obtained multiple monitors with the probability of error, and selecting the fuzzy optimization factor with the smallest probability of the fault as the optimal range of the number of monitoring units.
  • the method for optimizing the number of digital human-machine interface monitoring units of the present invention establishes a fuzzy optimization factor sequence optimized by the number of monitoring units by using fuzzy segmentation, and further selects a fuzzy optimization factor for monitoring by using a segmentation immune evolution method in a fuzzy optimization factor sequence.
  • the calculation of human error probability terminates the evolution process when monitoring the probability of human error becoming stable, reduces the number of iterations of evolution, reduces the amount of computational data, improves the computational efficiency, and thus optimizes optimal monitoring quickly and efficiently.
  • the range of the number of units to ensure the operator to obtain the efficiency and quality of the monitoring data improve the safety and reliability of the digital control system.
  • the preferred embodiment of the present invention selects a human-machine interface that can reflect the state of the plant in the event of a false alarm in the process of breaking the heat transfer tube of the steam generator for quantity optimization.
  • the test process also needs to partition the digitized human-machine interface first, because we will optimize the number of warning factors, parameter factors and information display factors in each functional partition.
  • These human-machine interfaces can be developed through the Visual Studio.net language platform.
  • the experimental process is as follows: First, set a range and the number of segments for each function block, that is, perform fuzzy optimization factor coding; the fuzzy segment takes the average segmentation value of the range interval, and the factor can be dynamically changed during the immune evolution process. Value range and fuzzy segment number; segmentation evolution of fuzzy optimization factor sequence by immune segmentation evolution method to extract fuzzy segment as fuzzy optimization factor; use fuzzy mapping factor to calculate affinity error rate for each fuzzy optimization factor to calculate Affinity failure rate that affects the results of the reliability model calculations; The extracted fuzzy optimization factor is monitored and transferred to Mrs.
  • Marco's reliability model calculation to calculate the probability of monitoring human error caused by the fuzzy optimization factor; to judge whether the probability of the error of the monitor is stable, and to monitor the probability of human error In the case of stability, the evolution is terminated; the probability of error is monitored and compared, and the fuzzy optimization factor with the smallest probability of error is extracted as the fuzzy segment with the optimal number of monitoring units.
  • the experimental data of the original interface and each post-immunization evolution interface are obtained through experiments, and then calculated according to the experimental data by the calculation method, and the error of the number of each functional block is obtained, and then each error is performed. Compare, so as to find a better number of function block blur, so as to achieve the purpose of optimization.
  • a digital human-machine interface monitoring unit quantity optimization system 100 includes: a fuzzy segmentation module 110, which performs fuzzy segmentation on the number of monitoring units of a digitized human-machine interface to be processed and establishes a plurality of fuzzy optimization factors.
  • a fuzzy optimization factor sequence a fuzzy optimization factor sequence
  • a segmentation extraction module 120 configured to perform segmentation immune evolution on the fuzzy optimization factor sequence to extract a fuzzy optimization factor from the fuzzy optimization factor sequence
  • a failure probability calculation module 130 configured to sequentially extract the blurred
  • the optimization factor is used to monitor the human error probability calculation to obtain the monitored human error probability of the extracted fuzzy optimization factor.
  • the evolution determination module 140 is configured to determine whether the probability of the human error caused by the error probability calculation module 130 is stable. That is, it is judged whether the difference between the maximum value and the minimum value of the monitored human error probability of the extracted fuzzy optimization factor is less than a predetermined threshold, and if so, the evolution is terminated, if otherwise, the segment extraction module 120 is returned to continue from the fuzzy optimization factor sequence.
  • the error probability calculation module 130 includes: an affinity error rate calculation unit 131, configured to perform an affinity error rate calculation on the extracted fuzzy optimization factor by using a mapping function, to calculate an impact monitoring and transfer of Mrs.
  • the error rate calculation unit 132 is configured to monitor the extracted fuzzy optimization factor by using the reliability model calculation to calculate the probability of monitoring the human error caused by the fuzzy optimization factor.
  • the above-described affinity error rate calculation unit 131 and error probability calculation unit 132 may be operated by a terminal processor.
  • the digital human-machine interface monitoring unit quantity optimization system of the present invention establishes a fuzzy optimization factor sequence for monitoring the quantity of the monitoring unit by using fuzzy segmentation, and further selects the fuzzy optimization factor in the fuzzy optimization factor sequence by the segmentation immune evolution method.
  • the evolution process is terminated when the probability of error is monitored, the number of iterations of evolution is reduced, the amount of computational data is reduced, and the computational efficiency is improved, so that the optimal optimization can be optimized quickly and efficiently.
  • the range of the number of monitoring units is to ensure that the operator obtains the efficiency and quality of the monitoring data and improves the safety and reliability of the digital control system.

Landscapes

  • Engineering & Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • Artificial Intelligence (AREA)
  • Software Systems (AREA)
  • General Physics & Mathematics (AREA)
  • Automation & Control Theory (AREA)
  • Health & Medical Sciences (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Evolutionary Computation (AREA)
  • Medical Informatics (AREA)
  • Mathematical Physics (AREA)
  • Fuzzy Systems (AREA)
  • Plasma & Fusion (AREA)
  • General Engineering & Computer Science (AREA)
  • High Energy & Nuclear Physics (AREA)
  • Testing And Monitoring For Control Systems (AREA)

Abstract

一种数字化人机界面监视单元数量优化方法及系统,该方法包括:对待处理的数字化人机界面的监视单元数量进行模糊分段以建立模糊优化因子序列;对模糊优化因子序列进行分段免疫进化以提取模糊优化因子,并分别对提取出的模糊优化因子进行监视人因失误概率计算;判断监视人因失误概率是否趋向稳定,若是则终止进化,若否则返回分段免疫进化的操作;对获得的多个监视人因失误概率进行比较,选取监视人因失误概率最小的模糊优化因子作为最佳的监视单元数量的取值范围。该方法及系统能够快速高效地优化出最佳的监视单元数量,以保证操作员获取监视数据的效率及质量,提高数字控制系统的安全可靠性。

Description

数字化人机界面监视单元数量优化方法及系统 技术领域 本发明涉及核电厂数字控制领域, 特别地, 涉及一种应用于核电厂数字控制系统 的数字化人机界面监视单元数量优化方法及系统。 背景技术 操作人员通过人机界面获取工厂当前运行状况的信息,将获得的信息,经判断、决策 之后执行相应的行为动作。经验表明,人机界面设计好坏对人的信息获取,判断等一系列 行为将带来影响。在数字控制系统发生的事故中,由人因引起的事故已占主要地位,人因 事故已成为当今核电站、 化工厂等事故的主要根源,例如,在系统失效中 ,60% -90 %归于 人误动作; 在 1995年美国统计中,大约 70%-90%事件与人有关。 数字化系统的新特点从根本上改变了操作人员对人机界面的认知行为、监视行为、 及应急行为等,而人因事故却是数字化主控室事故的根源,那么怎样改变这种状况呢? 一方面, 可以通过操作人员在整个数字化人机界面中的认知,监视等过程来提高操作人 员自身的适应能力、应变能力、判断能力、应急能力等等;另一方面,可以以人因可靠性 为基点对数字化人机界面进行优化,这样优化后的人机界面更能减少人因事故发生的 隐患。 现有的人机界面监视单元数量优化方法, 主要包括:
( 1 ) 有限微分法 这种方法是最原始的梯度方法,需要执行多次重复仿真,适用于瞬态和再生仿真。有 限微分方法为了获得更加可靠的估计值,需要更多次的运行来求得偏微分,使原本已经 很大的运行成本更高。
(2) 似然比法 似然比方法又称评分函数,仅需要单次仿真运行就可以得到梯度的估计值,其基本 思想是分析系统样本路径的概率测度对随机变量分布函数的依赖关系,通过测度变换 获得似然比,进而构造性能测度的估计量。 此方法也比较适合数值性问题; 该方法是通 过一些估计得到结果。 (3 ) 粒子群优化方法 使用 PSO 方法解决优化问题,把每个问题的潜在解定义为搜索空间中的一个粒子 01, 每个粒子可以用三元组(xi , vi , pi )表示,其中, xi表示粒子的当前位置; vi表示粒 子的当前速度; pi表示粒子本身搜索过的最好的位置 (个体经验)。 (4) 无约束间接法 无约束间接法是利用函数性态,通过微分或变分进行求优,主要有梯度法、 牛顿法、 变尺度法等。梯度法缺点是要求目标函数必须具有一阶偏导数,并需计算,迭代点离最优 点远时函数值下降快,越接近最优点收敛速度越慢。牛顿法缺点是要求目标函数必须有 一阶、 二阶偏导数及海森矩阵非奇异且正定或负定,需要计算一阶、 二阶偏导数及海森 矩阵的逆阵,程序复杂、计算量大。变尺度法缺点是需计算一阶偏导数,且由于舍入误差 和一维搜索的不精确等原因,数值稳定性仍不够理想,有时因计算误差引起变尺度矩阵 奇异而导致计算失败。
(5) 有约束直接法 该方法适用于仅含不等式约束的优化问题,具体有网络法、 随机方向搜索法以及复 合形法等。
(6) 神经网络方法 其实神经网络在优化方面应用比较少, 神经网络主要用来对输入的权重因子进行 修正及过程计算。 本研究主要用来对数字化人机界面的监视单元数量优化, 并未涉及 到对权重因子修正; 另一方面, 很难找到适合神经网络的激励函数, 因此神经网络不 适合本研究的监视单元数量优化问题。 现有的人机界面监视单元数量优化方法的数据运算量大、 迭代次数多, 无法满足 快速设计数字化人机界面监视单元数量的问题, 以使得操作人员能快速且准确地获得 监控信息, 降低因人机界面信息获取导致的人因事故的发生概率。 发明内容 本发明目的在于提供一种数字化人机界面监视单元数量优化方法, 以解决快速对 数字化人机界面监视单元的数量进行优化的技术问题。 本发明的另一目的在于提供一种数字化人机界面监视单元数量优化系统, 以解决 快速对数字化人机界面监视单元的数量进行优化的技术问题。 为实现上述目的, 本发明采用的技术方案如下: 一种数字化人机界面监视单元数量优化方法, 包括以下步骤: 对待处理的数字化人机界面的监视单元数量进行模糊分段以构成包括多个模糊优 化因子的模糊优化因子序列; 对所述模糊优化因子序列进行分段免疫进化以从所述模糊优化因子序列中提取模 糊优化因子; 依次对提取出的所述模糊优化因子进行监视人因失误概率计算, 以获取提取出的 所述模糊优化因子对应的监视人因失误概率; 判断所述监视人因失误概率是否趋向稳定, 即判断多个提取出的所述模糊优化因 子的监视人因失误概率的最大值与最小值的差值是否小于预定阈值, 若是则终止对所 述模糊优化因子序列的分段免疫进化, 若否则返回对所述模糊优化因子序列进行分段 免疫进化的步骤; 对获得的多个监视人因失误概率进行比较, 选取监视人因失误概率最小的模糊优 化因子作为最佳的监视单元数量的取值范围。 进一步地, 所述监视人因失误概率计算的步骤包括: 对提取出的所述模糊优化因子利用映射函数进行亲和力失误率运算, 以计算出影 响监视转移马尔可夫人因可靠性模型运算结果的亲和力失误率; 对提取出的所述模糊优化因子进行监视转移马尔可夫人因可靠性模型运算, 以计 算出该模糊优化因子对应的监视人因失误概率。 进一步地, 所述亲和力失误率运算采用的映射函数为:
1
P, = ( -) * η
1 + H 其中, 表示常量因子; Ρ 表示第 i个抗原因子中的第 j个模糊优化因子对抗体的 亲和力失误率; Hij表示抗体与抗原之间的亲和力。 根据本发明的另一方面, 一种数字化人机界面监视单元数量优化系统, 包括: 模糊分段模块, 用于对待处理的数字化人机界面的监视单元数量进行模糊分段以 构成包括多个模糊优化因子的模糊优化因子序列; 分段提取模块, 用于对所述模糊优化因子序列进行分段免疫进化以从所述模糊优 化因子序列中提取模糊优化因子; 失误概率计算模块, 依次对提取出的所述模糊优化因子进行监视人因失误概率计 算, 以获取提取出的所述模糊优化因子对应的监视人因失误概率; 进化判断模块, 用于判断所述监视人因失误概率是否趋向稳定, 即判断多个提取 出的所述模糊优化因子的监视人因失误概率的最大值与最小值的差值是否小于预定阈 值, 若是则终止对所述模糊优化因子序列的分段免疫进化, 若否则返回所述分段提取 模块以从所述模糊优化因子序列中提取所述模糊优化因子; 优化选取模块, 用于对所述失误概率计算模块获得的多个监视人因失误概率进行 比较, 选取监视人因失误概率最小的模糊优化因子作为最佳的监视单元数量的取值范 围。 进一步地, 所述失误概率计算模块包括: 亲和力失误率计算单元, 用于对提取出的所述模糊优化因子利用映射函数进行亲 和力失误率运算, 以计算出影响监视转移马尔可夫人因可靠性模型运算结果的亲和力 失误率; 失误概率计算单元, 用于对提取出的所述模糊优化因子进行监视转移马尔可夫人 因可靠性模型运算以计算出该模糊优化因子对应的监视人因失误概率。 本发明具有以下有益效果: 本发明数字化人机界面监视单元数量优化方法, 通过采用模糊分段建立用于监视 单元的数量优化的模糊优化因子序列, 并进一步通过分段免疫进化方法在模糊优化因 子序列中选取模糊优化因子以进行监视人因失误概率计算, 在监视人因失误概率趋向 稳定的情况下终止进化过程, 减少了进化的迭代次数, 精简了运算数据量, 提升了运 算效率, 从而能够快速高效地优化出最佳的监视单元数量的取值范围, 以保证操作员 获取监视数据的效率及质量, 提高数字控制系统的安全可靠性。 本发明数字化人机界面监视单元数量优化系统, 通过采用模糊分段建立用于监视 单元的数量优化的模糊优化因子序列, 并进一步通过分段免疫进化方法在模糊优化因 子序列中选取模糊优化因子以进行监视人因失误概率计算, 在监视人因失误概率趋向 稳定的情况下终止进化过程, 减少了进化的迭代次数, 精简了运算数据量, 提升了运 算效率, 从而能够快速高效地优化出最佳的监视单元数量的取值范围, 以保证操作员 获取监视数据的效率及质量, 提高数字控制系统的安全可靠性。 除了上面所描述的目的、特征和优点之外, 本发明还有其它的目的、特征和优点。 下面将参照图, 对本发明作进一步详细的说明。 附图说明 构成本申请的一部分的附图用来提供对本发明的进一步理解, 本发明的示意性实 施例及其说明用于解释本发明, 并不构成对本发明的不当限定。 在附图中: 图 1是本发明优选实施例数字化人机界面监视单元数量优化方法的步骤流程示意 图; 图 2是图 1中步骤 S30的流程示意图; 图 3是本发明优选实施例数字化人机界面监视单元数量优化系统的原理方框图; 以及 图 4是图 3中失误概率计算模块的原理方框图。 具体实施方式 以下结合附图对本发明的实施例进行详细说明, 但是本发明可以由权利要求限定 和覆盖的多种不同方式实施。 本发明用于解决数字化人机界面监视单元数量适合度问题, 以方便操作员通过人 机界面能够快速准确地获取监视信息, 以提高数字监控系统的安全可靠性。 参见图 1, 本发明的优选实施例提供了一种数字化人机界面监视单元数量优化方 法, 具体包括以下步骤: 步骤 S10: 对待处理的数字化人机界面的监视单元数量进行模糊分段并建立包括 多个模糊优化因子的模糊优化因子序列; 模糊理论主要用于研究现实世界中一些模糊不清的问题, 并使之清晰化。 本发明 利用模糊理论并结合并研究对象, 即数字化人机界面的监视单元数量, 将监视单元数 量进行模糊分段以构成模糊优化因子。 故模糊优化因子是对监视单元数量的数量段进 行编码的, 并不是对一个具体的数量值进行编码。 例如, 设定监视单元数量取值范围 为 1-20, 故模糊优化因子并不是 1-20之间的一个具体数值, 而只能是位于 1-20数量 范围内的数量段。 如将 1-20分为 [1,5]、 [6,10]、 [11,15]、 [16,20]四个模糊段, 则最后的 优化结果则为上述 4个模糊段中的一个模糊段。 步骤 S20: 对模糊优化因子序列进行分段免疫进化以从模糊优化因子序列中提取 出模糊优化因子; 传统优化方法在求解中小规模的优化问题时能够达到较高的速度和精度, 但实际 生活中大多数情况下面临的均是 NP完全问题, 由于求解复杂度较大, 传统的优化方 法已不能适应, 因为不可避免的出现 "维数灾"问题。在实践中有时为了避免这种情况, 经常采用强行中止,这必然导致大量的解空间被丢弃。 于是开始寻求新的算法, 例如: 模拟退火算法, 蚁群算法,粒子群算法, 鱼群算法等等。 本发明中, 对基于模糊分段的 模糊优化因子序列采用分段免疫进化方法搜索模糊优化因子, 对于搜索得到的模糊优 化因子进行监视人因失误概率计算。 具体的分段免疫进化方法可以分解为以下步骤: 先对模糊优化因子序列中的模糊段按模糊集由小到大的顺序进行排列, 分别对队 列 Ql, Q2进行初始化; 取模糊优化因子序列当中的中间段 d_ml进队列 Ql, 对队列 Q1进行出队操作, 这时中间段 d_ml将原模糊优化因子序列分为两个子序列; 将队列 Q1出队的模糊段按步长 n, 段数 m再分段, 依次把这些段进入队列 Q2, 队列 Q2 中的模糊段出列, 并依次对模糊段进行监视人因失误概率计算, 直至队 列 Q2为空; 重复返回至构建队列 Q1的步骤,具体为经提取中间段 d_ml后,余下的模糊段被 中间段 d_ml分为两个子序列, 分别取上述两个子序列的中间段进队列 Q1 ; 这样不断重复, 直至计算出的模糊段的监视人因失误概率趋于稳定, 即提取出的 模糊优化因子的监视人因失误概率中最大值与最小值的差值小于预设阈值。 步骤 S30: 依次对提取出的模糊优化因子进行监视人因失误概率计算, 以获取提 取出的模糊优化因子的监视人因失误概率; 在模糊优化因子序列中采用分段免疫进化方法提取出的模糊段即为模糊优化因 子, 参照图 2, 对模糊优化因子进行监视人因失误概率计算的步骤包括: 步骤 S31: 对提取出的模糊优化因子利用映射函数进行亲和力失误率运算, 以计 算出影响监视转移马尔可夫人因可靠性模型运算结果的亲和力失误率; 其中, 亲和力失误率运算的映射函数为:
1
=( -)*η (1)
1+H; 在公式 (1) 中, /;表示常量因子; Ρ 表示第 i个抗原因子中的第 j个模糊优化因 子对抗体的亲和力失误率; Hij表示抗体与抗原之间的亲和力, 当数字化人机界面监视 单元的优化因子取不同模糊集时, 得到的 j就会不同, 因此, 通过调整 Η 就可以达 到对人机界面因子的优化。 从公式(1)可知, 数字化人机界面参数数量设计的亲和力 越好, 亲和力失误率 Ρ 就越小, 而亲和力失误率 Ρ 会最终影响监视人因失误概率的 计算。 步骤 32: 对提取出的模糊优化因子进行监视转移马尔可夫人因可靠性模型运算以 计算出该模糊优化因子对应的监视人因失误概率。 通过执行步骤 S31和 S32, 即可对经分段进化提取出的各个模糊段进行监视人因 失误概率计算, 以得到各模糊段对应的监视人因失误概率。 其中, 监视转移马尔可夫人因可靠性模型的定义如下: p{MS; , H; , D; } = p(MS; , H; , D; I MSH } = p{MS; I Him, D; } + (p{Him } + ρ{0; }) (2) 在公式(2) 中, piMS^H DJ表示在人因、 决策行为影响下, 监视第 i个目标的 失误概率; piMS H DJMSi— J表示在人的影响因子及决策行为影响下, 从第 i-1 个 目标转移到第 i个目标的失误率; p{MSilHim,Di}表示在人的影响因子及决策行为影 响下, 监视第 i个目标的失误率; p{Him}表示监视中, 人的影响因子所造成的监视失 误率, ρ{ }表示监视中, 决策行为造成的监视失误率。 对 p{Him}的计算考虑了模糊 数量优化因子、 人的影响因子及监视时间, p{Him}的计算式中包含了亲和力失误率 Ρϋ, 故在应用监视转移马尔可夫人因可靠性模型计算监视人因失误概率需要先计算出 亲和力失误率 Pij 。 步骤 S40: 判断监视人因失误概率是否趋向稳定, 即判断取出的模糊优化因子的 监视人因失误概率中最大值与最小值的差值是否小于预定阈值, 若是则终止对模糊优 化因子序列进行分段免疫进化, 执行步骤 S50; 若否, 则返回步骤 S20继续对模糊优 化因子序列进行分段免疫进化, 以提取深度进化的模糊段作为模糊优化因子。 步骤 S50: 对获得的多个监视人因失误概率进行比较, 选取监视人因失误概率最 小的模糊优化因子作为最佳的监视单元数量的取值范围。 本发明数字化人机界面监视单元数量优化方法, 通过采用模糊分段建立监视单元 的数量优化的模糊优化因子序列, 并进一步通过分段免疫进化方法在模糊优化因子序 列中选取模糊优化因子以进行监视人因失误概率计算, 在监视人因失误概率趋向稳定 的情况下终止进化过程, 减少了进化的迭代次数, 精简了运算数据量, 提升了运算效 率, 从而能够快速高效地优化出最佳的监视单元数量的取值范围, 以保证操作员获取 监视数据的效率及质量, 提高数字控制系统的安全可靠性。 本发明优选实施例选取在蒸汽发生器传热管断裂过程中的误安注事件中能总体反 映工厂状态的人机界面来进行数量优化。 该试验过程先也需要对该数字化人机界面进 行分区, 因为我们将对每个功能分区中的警告因子, 参数因子及信息显示因子的数量 进行优化。 我们以误安注主界面为例进行考虑。 实验用到的模拟界面共 46 个, 其中 35个辅助界面是该情景下相关的人机界面, 另外 11个界面是原始界面及各功能块参 数数量变量过程中的演化界面,是优化的对象。这些人机界面均可通过 Visual studio.net 语言平台开发出来, 由于界面较多, 这里不予列出。 实验过程如下: 先对每个功能块的因子设定一个范围及分段数, 即进行模糊优化因子编码; 模糊 段取范围区间的平均分段值, 在免疫进化过程中, 可以动态改变因子的取值范围和模 糊分段数; 利用免疫分段进化方法对模糊优化因子序列进行分段进化, 以提取模糊段作为模 糊优化因子; 对各模糊优化因子利用映射函数进行亲和力失误率运算, 以计算出影响监视转移 马尔可夫人因可靠性模型运算结果的亲和力失误率; 对提取出的模糊优化因子进行监视转移马尔可夫人因可靠性模型运算以计算出该 模糊优化因子对应的监视人因失误概率; 判断监视人因失误概率是否趋向稳定, 在监视人因失误概率趋向稳定的情况下, 终止进化; 比较监视人因失误概率, 提取监视人因失误概率最小的模糊优化因子作为监视单 元数量最优的模糊段。 本实施例通过实验获得了原始界面及每次分段免疫进化后界面的有关实验数据, 再根据实验数据通过计算方法进行计算, 得到每种功能块数量的失误情况, 再把每种 失误情况进行比较, 从而找出比较好的功能块模糊数量, 从而达到优化的目的。 通过 分析我们得到了蒸汽发生器传热管断裂过程中误发安注功能块几个对象的数量优化的 最好结果。 对该试验获得的数据进行分析, 可以得到以下两点: 第一, 分段免疫进化方法在 达到最优解时的平均进化次数比顺序进化次数少得多; 第二, 本发明分段免疫进化方 法具有收敛快, 稳定性高, 灵敏度好等优点。 参照图 3, 一种数字化人机界面监视单元数量优化系统 100, 包括: 模糊分段模块 110, 用于对待处理的数字化人机界面的监视单元数量进行模糊分 段并建立包括多个模糊优化因子的模糊优化因子序列; 分段提取模块 120, 用于对模糊优化因子序列进行分段免疫进化以从模糊优化因 子序列中提取模糊优化因子; 失误概率计算模块 130, 用于依次对提取出的模糊优化因子进行监视人因失误概 率计算, 以获取提取出的模糊优化因子的监视人因失误概率; 进化判断模块 140, 用于判断失误概率计算模块 130计算出的监视人因失误概率 是否趋向稳定, 即判断提取出的模糊优化因子的监视人因失误概率中最大值与最小值 的差值是否小于预定阈值, 若是则终止进化, 若否则返回分段提取模块 120, 以继续 从模糊优化因子序列中选取模糊优化因子; 优化选取模块 150, 用于对失误概率计算模块 130获得的多个监视人因失误概率 进行比较, 选取监视人因失误概率最小的模糊优化因子作为最佳的监视单元数量的取 值范围。 可以通过终端处理器来运行上述模糊分段模块 110、分段提取模块 120、进化判断 模块 140和优化选取模块 150。 进一步地, 参照图 4, 失误概率计算模块 130包括: 亲和力失误率计算单元 131, 用于对提取出的模糊优化因子利用映射函数进行亲 和力失误率运算, 以计算出影响监视转移马尔可夫人因可靠性模型运算结果的亲和力 失误率; 失误概率计算单元 132, 用于对提取出的模糊优化因子进行监视转移马尔可夫人 因可靠性模型运算以计算出该模糊优化因子对应的监视人因失误概率。 可以通过终端处理器来运行上述亲和力失误率计算单元 131和失误概率计算单元 132。 本发明数字化人机界面监视单元数量优化系统, 通过采用模糊分段建立用于监视 单元的数量优化的模糊优化因子序列, 并进一步通过分段免疫进化方法在模糊优化因 子序列中选取模糊优化因子以进行监视人因失误概率计算, 在监视人因失误概率趋向 稳定的情况下终止进化过程, 减少了进化的迭代次数, 精简了运算数据量, 提升了运 算效率, 从而能够快速高效地优化出最佳的监视单元数量的取值范围, 以保证操作员 获取监视数据的效率及质量, 提高数字控制系统的安全可靠性。 以上所述仅为本发明的优选实施例而已, 并不用于限制本发明, 对于本领域的技 术人员来说, 本发明可以有各种更改和变化。 凡在本发明的精神和原则之内, 所作的 任何修改、 等同替换、 改进等, 均应包含在本发明的保护范围之内。

Claims

权 利 要 求 书
1. 一种数字化人机界面监视单元数量优化方法, 其特征在于, 包括以下步骤: 对待处理的数字化人机界面的监视单元数量进行模糊分段, 以构成包括多 个模糊优化因子的模糊优化因子序列;
对所述模糊优化因子序列进行分段免疫进化, 以从所述模糊优化因子序列 中提取模糊优化因子;
依次对提取出的所述模糊优化因子进行监视人因失误概率计算, 以获取提 取出的所述模糊优化因子对应的监视人因失误概率;
判断多个提取出的所述模糊优化因子的监视人因失误概率的最大值与最小 值的差值是否小于预定阈值, 若是则终止对所述模糊优化因子序列的分段免疫 进化, 若否, 则返回对所述模糊优化因子序列进行分段免疫进化的步骤; 对获得的多个监视人因失误概率进行比较, 选取监视人因失误概率最小的 模糊优化因子作为最佳的监视单元数量的取值范围。
2. 根据权利要求 1所述的数字化人机界面监视单元数量优化方法, 其特征在于, 对所述模糊优化因子序列进行分段免疫进化包括如下步骤: 对所述模糊优化因子序列中的模糊段按模糊集由小到大的顺序进行排列, 分别对队列 Ql, Q2进行初始化;
取所述模糊优化因子序列当中的中间段 d_ml进所述队列 Ql,对所述队列 Q1进行出队操作,使得所述中间段(1_1^将所述模糊优化因子序列分为两个子 序列;
将所述队列 Q1出队的模糊段按步长 n, 段数 m再分段, 依次把这些段进 入所述队列 Q2;
所述队列 Q2中的模糊段出列, 并依次对模糊段进行监视人因失误概率计 算, 直至队列 Q2为空; 重复返回至构建所述队列 Q1 的步骤进行循环, 直至提取出的所述模糊优 化因子的监视人因失误概率中最大值与最小值的差值小于所述预设阈值。
3. 根据权利要求 1所述的数字化人机界面监视单元数量优化方法, 其特征在于, 所述监视人因失误概率计算的步骤包括:
对提取出的所述模糊优化因子利用映射函数进行亲和力失误率运算, 以计 算出影响监视转移马尔可夫人因可靠性模型运算结果的亲和力失误率;
对提取出的所述模糊优化因子进行监视转移马尔可夫人因可靠性模型运 算, 以计算出该模糊优化因子对应的监视人因失误概率。
4. 根据权利要求 3所述的数字化人机界面监视单元数量优化方法, 其特征在于, 所述亲和力失误率运算采用的映射函数为:
Figure imgf000014_0001
其中, /;表示常量因子; 表示第 i个抗原因子中的第 j个模糊优化因子对 抗体的亲和力失误率; ί¾表示所述抗体与抗原之间的亲和力。
5. 根据权利要求 3所述的数字化人机界面监视单元数量优化方法, 其特征在于, 通过如下计算公式来获取所述监视转移马尔可夫人因可靠性模型: p{MSi,Hi,Di}= piMS^H^D; I MS;^ } = p{MS; I Him,Di} + (p{Him}+ plDJ) , 其中, PiMS P^DJ表示在人因、 决策行为影响下, 监视第 i个目标的失误 概率; P M^'^'D^MSw}表示在人的影响因子及决策行为影响下, 从第 i-1 个目标转移到第 i个目标的失误率; p{MSl lH 'Dl}表示在人的影响因子及决 策行为影响下, 监视第 i个目标的失误率; p{H }表示监视中, 人的影响因子 所造成的监视失误率, p{Dl}表示监视中, 决策行为造成的监视失误率。
6. 一种数字化人机界面监视单元数量优化系统, 其特征在于, 包括:
模糊分段模块, 用于对待处理的数字化人机界面的监视单元数量进行模糊 分段, 以构成包括多个模糊优化因子的模糊优化因子序列;
分段提取模块, 用于对所述模糊优化因子序列进行分段免疫进化, 以从所 述模糊优化因子序列中提取模糊优化因子;
失误概率计算模块, 依次对提取出的所述模糊优化因子进行监视人因失误 概率计算, 以获取提取出的所述模糊优化因子对应的监视人因失误概率; 进化判断模块, 用于判断多个提取出的所述模糊优化因子的监视人因失误 概率的最大值与最小值的差值是否小于预定阈值, 若是则终止对所述模糊优化 因子序列的分段免疫进化, 若否则返回所述分段提取模块以从所述模糊优化因 子序列中提取所述模糊优化因子;
优化选取模块, 用于对所述失误概率计算模块获得的多个监视人因失误概 率进行比较, 选取监视人因失误概率最小的模糊优化因子作为最佳的监视单元 数量的取值范围。
7. 根据权利要求 6所述的数字化人机界面监视单元数量优化系统, 其特征在于, 所述分段提取模块包括:
第一排序模块, 用于对所述模糊优化因子序列中的模糊段按模糊集由小到 大的顺序进行排列, 分别对队列 Ql, Q2进行初始化;
第一队列处理模块, 用于取所述模糊优化因子序列当中的中间段(1_1^进 所述队列 Ql, 对所述队列 Q1进行出队操作, 使得所述中间段 d_ml将所述模 糊优化因子序列分为两个子序列;
分段模块, 用于将所述队列 Q1出队的模糊段按步长 n, 段数 m再分段, 依次把这些段进入所述队列 Q2;
第二队列处理模块, 用于所述队列 Q2中的模糊段出列, 并依次对模糊段 进行监视人因失误概率计算, 直至队列 Q2为空;
循环模块, 用于重复返回至构建所述队列 Q1 的步骤进行循环, 直至提取 出的所述模糊优化因子的监视人因失误概率中最大值与最小值的差值小于所述 预设阈值。
8. 根据权利要求 6所述的数字化人机界面监视单元数量优化系统, 其特征在于, 所述失误概率计算模块包括:
亲和力失误率计算单元, 用于对提取出的所述模糊优化因子利用映射函数 进行亲和力失误率运算, 以计算出影响监视转移马尔可夫人因可靠性模型运算 结果的亲和力失误率;
失误概率计算单元, 用于对提取出的所述模糊优化因子进行监视转移马尔 可夫人因可靠性模型运算以计算出该模糊优化因子对应的监视人因失误概率。
9. 根据权利要求 8所述的数字化人机界面监视单元数量优化装置, 其特征在于, 所述亲和力失误率计算单元包括: 第一计算模块, 用于提供所述亲和力失误率运算采用的映射函数, 所述映 射函数为:
Figure imgf000016_0001
其中, ;;表示常量因子; 表示第 i个抗原因子中的第 j个模糊优化因子对 抗体的亲和力失误率; j表示所述抗体与抗原之间的亲和力。
10. 根据权利要求 8所述的数字化人机界面监视单元数量优化系统, 其特征在于, 所述失误概率计算单元包括:
第二计算模块, 用于通过如下计算公式来获取所述监视转移马尔可夫人因 可靠性模型:
p{MSi,Hi,Di}= piMS^H^D; I MS;^ } = p{MS; I Him,Di} + (p{Him}+ plDJ) , 其中, PiMS P^DJ表示在人因、 决策行为影响下, 监视第 i个目标的失误 概率; P M^'^'D^MSw}表示在人的影响因子及决策行为影响下, 从第 i-1 个目标转移到第 i个目标的失误率; p{MSl lH 'Dl}表示在人的影响因子及决 策行为影响下, 监视第 i个目标的失误率; p{H }表示监视中, 人的影响因子 所造成的监视失误率, p{Dl}表示监视中, 决策行为造成的监视失误率。
PCT/CN2014/075824 2013-04-23 2014-04-21 数字化人机界面监视单元数量优化方法及系统 WO2014173271A1 (zh)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN201310141760.XA CN103197564B (zh) 2013-04-23 2013-04-23 数字化人机界面监视单元数量优化方法及系统
CN201310141760.X 2013-04-23

Publications (1)

Publication Number Publication Date
WO2014173271A1 true WO2014173271A1 (zh) 2014-10-30

Family

ID=48720230

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2014/075824 WO2014173271A1 (zh) 2013-04-23 2014-04-21 数字化人机界面监视单元数量优化方法及系统

Country Status (2)

Country Link
CN (1) CN103197564B (zh)
WO (1) WO2014173271A1 (zh)

Families Citing this family (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN103197564B (zh) * 2013-04-23 2016-01-20 湖南工学院 数字化人机界面监视单元数量优化方法及系统
CN103970018A (zh) * 2014-05-07 2014-08-06 湖南工学院 获取人机交互复杂度数据的方法和装置
CN109445484B (zh) * 2018-12-05 2020-09-15 长安大学 一种基于猫群优化和免疫模糊pid的孵化室温度控制方法
CN110400646B (zh) * 2019-04-28 2024-07-09 北京广利核系统工程有限公司 应用于核电站安全级dcs的人机交互装置、系统及方法

Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US5319579A (en) * 1990-02-05 1994-06-07 Westinghouse Electric Corp. Simulation of human performance of a process operation procedure
US5519740A (en) * 1992-07-06 1996-05-21 General Electric Company Reactor control having transient mitigation system
CN101840694A (zh) * 2009-09-15 2010-09-22 中广核工程有限公司 核电厂数字化规程人机界面的显示装置及其显示控制方法
CN102393834A (zh) * 2011-10-21 2012-03-28 中广核工程有限公司 一种核电站数字化人机界面的测试方法和系统
CN103197564A (zh) * 2013-04-23 2013-07-10 湖南工学院 数字化人机界面监视单元数量优化方法及系统

Family Cites Families (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP2004206438A (ja) * 2002-12-25 2004-07-22 Mitsubishi Electric Corp ヒューマンエラー評価支援装置及びプログラム
KR101033974B1 (ko) * 2009-12-11 2011-05-11 한국수력원자력 주식회사 소프트 제어기 유형에 따른 인적오류 모드 및 psf 생성 장치
CN102915512A (zh) * 2012-09-25 2013-02-06 中国核电工程有限公司 一种基于数字化人机界面的核电站安全操作评价方法

Patent Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US5319579A (en) * 1990-02-05 1994-06-07 Westinghouse Electric Corp. Simulation of human performance of a process operation procedure
US5519740A (en) * 1992-07-06 1996-05-21 General Electric Company Reactor control having transient mitigation system
CN101840694A (zh) * 2009-09-15 2010-09-22 中广核工程有限公司 核电厂数字化规程人机界面的显示装置及其显示控制方法
CN102393834A (zh) * 2011-10-21 2012-03-28 中广核工程有限公司 一种核电站数字化人机界面的测试方法和系统
CN103197564A (zh) * 2013-04-23 2013-07-10 湖南工学院 数字化人机界面监视单元数量优化方法及系统

Non-Patent Citations (1)

* Cited by examiner, † Cited by third party
Title
WANG, ZHIFANG ET AL.: "Human Factor Engineering Analysis for Computerized Human Machine Interface Design Issues", NUCLEAR SCIENCE AND ENGINEERING, vol. 30, no. 4, 31 December 2010 (2010-12-31), pages 366 - 371 *

Also Published As

Publication number Publication date
CN103197564A (zh) 2013-07-10
CN103197564B (zh) 2016-01-20

Similar Documents

Publication Publication Date Title
CN111124840B (zh) 业务运维中告警的预测方法、装置与电子设备
CN109492822B (zh) 空气污染物浓度时空域关联预测方法
CN108764072B (zh) 一种基于多尺度融合的血细胞亚型图像分类方法
KR101822829B1 (ko) 머신러닝을 이용한 차단기 부분방전 유형진단 및 노이즈 판별 방법
CN108809948B (zh) 一种基于深度学习的异常网络连接检测方法
CN109033513B (zh) 电力变压器故障诊断方法与电力变压器故障诊断装置
CN112910859B (zh) 基于c5.0决策树和时序分析的物联网设备监测预警方法
CN106021771A (zh) 一种故障诊断方法及装置
CN105354198B (zh) 一种数据处理方法及装置
CN113762486B (zh) 换流阀故障诊断模型的构建方法、装置和计算机设备
CN112966714B (zh) 一种边缘时序数据异常检测和网络可编程控制方法
CN107247450A (zh) 基于贝叶斯网络的断路器故障诊断方法
WO2014173271A1 (zh) 数字化人机界面监视单元数量优化方法及系统
CN110113327A (zh) 一种检测dga域名的方法及装置
CN111126820A (zh) 反窃电方法及系统
CN110851654A (zh) 基于张量化数据降维的工业设备故障检测分类方法
CN112215279B (zh) 一种基于免疫rbf神经网络的电网故障诊断方法
CN111753751A (zh) 一种改进烟花算法的风机故障智能诊断方法
CN114611372A (zh) 一种基于物联网边缘计算的工业设备健康预测方法
CN117041017A (zh) 数据中心的智能运维管理方法及系统
CN115329663A (zh) 处理电力负荷监测稀疏数据的关键特征选择方法和装置
CN113541985A (zh) 物联网故障诊断方法、模型的训练方法及相关装置
CN111400964B (zh) 一种故障发生时间预测方法及装置
CN112203311B (zh) 网元异常诊断方法、装置、设备及计算机存储介质
CN116400168A (zh) 一种基于深度特征聚类的电网故障诊断方法及系统

Legal Events

Date Code Title Description
NENP Non-entry into the national phase

Ref country code: DE

121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 14788077

Country of ref document: EP

Kind code of ref document: A1

122 Ep: pct application non-entry in european phase

Ref document number: 14788077

Country of ref document: EP

Kind code of ref document: A1