CN108182337B - 一种基于ScMiUKFNN算法的天然气净化工艺建模方法 - Google Patents

一种基于ScMiUKFNN算法的天然气净化工艺建模方法 Download PDF

Info

Publication number
CN108182337B
CN108182337B CN201810200655.1A CN201810200655A CN108182337B CN 108182337 B CN108182337 B CN 108182337B CN 201810200655 A CN201810200655 A CN 201810200655A CN 108182337 B CN108182337 B CN 108182337B
Authority
CN
China
Prior art keywords
neural network
network model
state
covariance
time
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.)
Active
Application number
CN201810200655.1A
Other languages
English (en)
Other versions
CN108182337A (zh
Inventor
辜小花
王甜
唐海红
张堃
宋鸿飞
张兴
侯松
裴仰军
李太福
邱奎
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Chongqing University of Science and Technology
Original Assignee
Chongqing University of Science and Technology
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 Chongqing University of Science and Technology filed Critical Chongqing University of Science and Technology
Priority to CN201810200655.1A priority Critical patent/CN108182337B/zh
Publication of CN108182337A publication Critical patent/CN108182337A/zh
Application granted granted Critical
Publication of CN108182337B publication Critical patent/CN108182337B/zh
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F30/00—Computer-aided design [CAD]
    • G06F30/20—Design optimisation, verification or simulation
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N3/00—Computing arrangements based on biological models
    • G06N3/02—Neural networks
    • G06N3/04—Architecture, e.g. interconnection topology

Landscapes

  • Engineering & Computer Science (AREA)
  • Theoretical Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • Evolutionary Computation (AREA)
  • General Physics & Mathematics (AREA)
  • General Engineering & Computer Science (AREA)
  • General Health & Medical Sciences (AREA)
  • Artificial Intelligence (AREA)
  • Data Mining & Analysis (AREA)
  • Biophysics (AREA)
  • Biomedical Technology (AREA)
  • Molecular Biology (AREA)
  • Computing Systems (AREA)
  • Computational Linguistics (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Mathematical Physics (AREA)
  • Software Systems (AREA)
  • Health & Medical Sciences (AREA)
  • Computer Hardware Design (AREA)
  • Geometry (AREA)
  • Gas Separation By Absorption (AREA)
  • Treating Waste Gases (AREA)

Abstract

本发明公开了一种基于ScMiUKFNN算法的天然气净化工艺建模方法，包括以下步骤：步骤S1：选择影响脱硫效率的工艺参数和脱硫单元的性能指标；步骤S2：采集预设时间的所述工艺参数和所述性能指标的数据；步骤S3：形成归一化样本集，取所述归一化样本集中一部分作为训练样本，剩余部分作为测试样本；步骤S4：基于训练样本构建神经网络模型和所述神经网络模型的初始状态变量；步骤S5：利用ScMiUKFNN算法估计所述神经网络模型的最优状态变量；步骤S6：获得训练样本更新后的神经网络模型；步骤S7：得到预测结果，将预测结果与所述测试样本中的实际输出进行比较，如果比较结果小于预设误差值，神经网络模型有效；否则重复上述步骤至比较结果小于预设误差值。

Description

一种基于ScMiUKFNN算法的天然气净化工艺建模方法
技术领域
本发明涉及高含硫天然气净化技术领域，更为具体地，涉及一种基于ScMiUKFNN算法的天然气净化工艺建模方法。
背景技术
随着清洁能源需求的快速增长，天然气的需求在逐渐增加。然而，高含硫气体(high-sulfur gas,HSG)含酸性气体比一般天然气多十倍，在中国的气藏中占有相当大的比例。由于具有毒性和腐蚀性，高含硫气体不能直接使用，必须去除硫化氢(H2S)和二氧化碳(CO2)，称为高含硫气体脱硫过程。此外，随着酸性气体吸收剂溶液循环的增加，这一过程的能量消耗和生产成本急剧增加。因此，降低能源消耗和运行成本，提高企业的经济效益，提高市场竞争力，是一个急待解决的问题。
发明内容
本发明的目的在于克服现有技术的不足，提供一种基于ScMiUKFNN算法的天然气净化工艺建模方法
本发明的目的是这样实现的：
一种基于ScMiUKFNN算法的天然气净化工艺建模方法，本方法包括以下步骤：
步骤S1：选择影响脱硫效率的工艺参数和脱硫单元的性能指标；
步骤S2：采集预设时间的所述工艺参数和所述性能指标的数据，剔除误差样本后形成样本集[X，Y]；
步骤S3：对样本集[X，Y]进行归一化，形成归一化样本集
Figure BDA0001594406700000021
取所述归一化样本集
Figure BDA0001594406700000022
中一部分样本作为训练样本，剩余部分的样本作为测试样本；
步骤S4：基于所述训练样本构建神经网络模型和所述神经网络模型的初始状态变量θk，以及，将所述训练样本中的
Figure BDA0001594406700000023
作为所述神经网络模型的输入，将所述训练样本中的
Figure BDA0001594406700000024
作为所述神经网络模型的输出；
所述神经网络模型为：
Figure BDA0001594406700000025
Figure BDA0001594406700000026
其中，
Figure BDA0001594406700000027
为所述训练样本的矢量样本值，并作为所述神经网络模型的输入；zj作为所述神经网络wij模型的隐含层输出；
Figure BDA0001594406700000028
作为所述神经网络模型的输出层输出；wij为神经网络模型的输入层到隐含层的神经元的连接权值；
Figure BDA0001594406700000029
为神经网络模型的输入层到隐含层的神经元的阈值；vjd为所述神经网络模型的隐含层到输出层的神经元的连接权值，
Figure BDA00015944067000000210
为所述神经网络模型的隐含层到输出层的神经元的阈值，i＝1，2，…，m；m为神经网络模型的输入层的神经元的数量，s为神经网络模型的隐含层的神经元的数量，h为神经网络模型的输出层的神经元的数量；
应用于神经网络模型各层神经元的非线性激活函数为：
Figure BDA00015944067000000211
fo(x)＝x (4)
所述初始状态变量为：
Figure BDA0001594406700000031
步骤S5：利用ScMiUKFNN算法估计所述神经网络模型的最优状态变量；
步骤S6：将所述最优状态变量作为所述神经网络模型的wij、vjd、
Figure BDA0001594406700000032
和
Figure BDA0001594406700000033
对公式(1)和公式(2)进行更新，获得训练样本更新后的神经网络模型；
步骤S7：将所述测试样本中的
Figure BDA00015944067000000314
输入到更新后的神经网络模型，得到预测结果，将所述预测结果与所述测试样本中的实际输出
Figure BDA00015944067000000315
进行比较，如果比较结果小于预设误差值，所构建的神经网络模型有效；否则重复上述步骤S1-S7，直至所述比较结果小于所述预设误差值为止。
优选地，所述步骤S5包括：
步骤S51：在建立的所述神经网络模型中，将参数向量视为算法所需的状态方程，网络输出视为量测方程：
Figure BDA0001594406700000034
其中，
Figure BDA0001594406700000035
为神经网络模型的输入，
Figure BDA0001594406700000036
为神经网络模型的输出，
Figure BDA0001594406700000037
是参数化的非线性函数，ηk是过程噪声，μk是测量噪声；
对状态方程和量测方程进行初始化，计算状态变量估计以及其协方差：
Figure BDA0001594406700000038
Figure BDA0001594406700000039
其中：
Figure BDA00015944067000000310
是状态值，Pk是协方差矩阵；
步骤S52：引入一个最小的sigma集合，运用减少Sigma点集方法对所述初始状态变量θk进行Sigma采样，获得n+1个采样点以及权重系数，以减少计算复杂度；随机变量
Figure BDA00015944067000000311
具有均值
Figure BDA00015944067000000312
和协方差矩阵PXX＞0，则：
Figure BDA00015944067000000313
Figure BDA0001594406700000041
Wweight=[W ωn+1] (9)
其中：
Figure BDA0001594406700000042
Figure BDA0001594406700000043
“：＝”为赋值号；
步骤S53：状态更新，通过离散时间非线性系统的状态方程将每个采样点的k时刻的最优状态变量的状态估计变换为k+1时刻的状态变量的状态估计
Figure BDA0001594406700000044
并通过合并k+1时刻的状态估计
Figure BDA0001594406700000045
的向量，获得k+1时刻的状态变量的状态先验估计
Figure BDA0001594406700000046
和协方差
Figure BDA0001594406700000047
其中，所述状态估计
Figure BDA0001594406700000048
为：
Figure BDA0001594406700000049
其中，β为放缩因子，f为线性方程；
所述状态先验估计
Figure BDA00015944067000000410
为：
Figure BDA00015944067000000411
所述状态变量的协方差
Figure BDA00015944067000000412
为：
Figure BDA00015944067000000413
步骤S54：量测更新，通过离散时间非线性系统的量测方程建立k时刻的状态变量的状态估计
Figure BDA00015944067000000414
和k时刻的量测预测估计
Figure BDA00015944067000000415
之间的联系以完成量测预测，并估计k时刻的量测预测的协方差
Figure BDA0001594406700000051
以及k时刻的状态变量和量测预测之间的协方差
Figure BDA0001594406700000052
其中，所述量测估计
Figure BDA0001594406700000053
为：
Figure BDA0001594406700000054
所述k时刻的量测预测的均值
Figure BDA0001594406700000055
为：
Figure BDA0001594406700000056
所述k时刻的量测预测的协方差
Figure BDA0001594406700000057
为：
Figure BDA0001594406700000058
所述k时刻的状态变量和量测预测之间的协方差
Figure BDA0001594406700000059
为：
Figure BDA00015944067000000510
步骤S55：通过建立协方差
Figure BDA00015944067000000511
和协方差
Figure BDA00015944067000000512
的关系，更新k时刻的状态变量的状态估计和协方差；
所述协方差之间的关系是：
Figure BDA00015944067000000513
通过上述关系对k+1时刻的状态变量的状态估计和协方差进行修正：
Figure BDA00015944067000000514
Figure BDA00015944067000000515
步骤S56：将获得的修正后k+1时刻的状态变量
Figure BDA00015944067000000516
重组神经网络模型，并计算此时神经网络模型的预测输出与实际输出之间的误差，如果小于既设精度要求，则输出所述神经网络模型的最优状态变量
Figure BDA00015944067000000517
反之，重新进入步骤1。
优选地，所述工艺参数包括进入尾气吸收塔的贫胺液流量、进入二级吸收塔的贫胺液流量、原料气处理量、尾气单元返回脱硫单元的半富胺液流量、一级吸收塔胺液入塔温度、二级吸收塔胺液入塔温度、闪蒸罐压力、一个重沸器的蒸汽消耗量、另一个重沸器的蒸汽消耗量和蒸汽预热器的蒸汽消耗量；脱硫单元的性能指标包括净化气中H2S和CO2的浓度。
优选地，步骤S3中，取所述归一化样本集
Figure BDA0001594406700000061
中前80％的样本作为训练样本，而剩余的20％样本作为测试样本。
由于采用了上述技术方案，本发明相对于现有技术能够节能降耗，提高产率和气体加工经济效益。
附图说明
图1a、图1b为比较预测结果和运行数据图；
图2a、图2b为对模型精度的比较图；
图3a、图3b为比较预测结果和运行数据图。
具体实施方式
名词解释
ScMiUKFNN：Scaled Minimum Unscented Kalman Filter Neural Network，基于比例放缩的减少采样无迹卡尔曼滤波神经网络。
本发明提供的基于ScMiUKFNN算法的工业过程建模方法，包括：
步骤S1：选择影响脱硫效率的工艺参数和脱硫单元的性能指标；其中，工艺参数包括进入尾气吸收塔贫的胺液流量、进入二级吸收塔的贫胺液流量、原料气处理量、尾气单元返回脱硫单元的半富胺液流量、一级吸收塔胺液入塔温度、二级吸收塔胺液入塔温度、闪蒸罐压力、一个重沸器的蒸汽消耗量、另一个重沸器的蒸汽消耗量和蒸汽预热器的蒸汽消耗量；脱硫单元的性能指标包括净化气中H2S和CO2的浓度。参数列表如表1所示：
表1
Figure BDA0001594406700000072
步骤S2：采集预设时间的所述工艺参数和所述性能指标的数据，剔除误差样本后形成样本集[X，Y]；样本集[X，Y]如下表2所示：
表2
Figure BDA0001594406700000071
Figure BDA0001594406700000081
步骤S3：对样本集[X，Y]进行归一化，形成归一化样本集
Figure BDA0001594406700000082
取所述归一化样本集
Figure BDA0001594406700000083
中前80％的样本作为训练样本，而剩余的20％样本作为测试样本；
步骤S4：基于所述训练样本构建神经网络模型和所述神经网络模型的初始状态变量θk，以及，将所述训练样本中的
Figure BDA0001594406700000084
作为所述神经网络模型的输入，将所述训练样本中的
Figure BDA0001594406700000085
作为所述神经网络模型的输出；
其中，所述神经网络模型为：
Figure BDA0001594406700000086
Figure BDA0001594406700000087
其中，
Figure BDA0001594406700000088
为所述训练样本的矢量样本值，并作为所述神经网络模型的输入；zj作为所述神经网络隐含层输出；
Figure BDA0001594406700000089
作为所述神经网络输出层输出；wij为网络输入层到隐含层的神经元的连接权值；
Figure BDA0001594406700000091
为网络输入层到所述隐含层的神经元的阈值；vjd为所述隐含层到网络输出层的神经元的连接权值，
Figure BDA0001594406700000092
为所述隐含层到所述网络输出层的神经元的阈值，其中，i＝1，2，…，m；m为所述网络输入层的神经元的数量，s为所述网络隐含层的神经元的数量，h为所述网络输出层的神经元的数量；
其中，应用于各层神经元的非线性激活函数为：
Figure BDA0001594406700000093
fo(x)＝x (4)
所述初始状态变量为：
Figure BDA0001594406700000094
步骤S5：利用ScMiUKFNN算法估计所述神经网络模型的最优状态变量；
本发明利用ScMiUKFNN算法估计神经网络模型的状态变量，以达到连接权值、阈值的不断调整，直到满足要求。将得到的最优状态变量的状态估计作为上述所建立神经网络模型的连接权值、阈值。需要说明的是，该连接权值、阈值为通过ScMiUKFNN算法调整后的连接权值、阈值，也是上述所建立的神经网络模型的全部连接权值与阈值，包括wij、vjd、
Figure BDA0001594406700000095
和
Figure BDA0001594406700000096
利用ScMiUKFNN算法估计神经网络模型的最优状态变量的过程包括：
步骤S51：在建立的所述神经网络模型中，将参数向量视为算法所需的状态方程，网络输出可视为量测方程：
Figure BDA0001594406700000097
其中，
Figure BDA0001594406700000098
为神经网络模型的输入，
Figure BDA0001594406700000099
为神经网络模型的输出，
Figure BDA00015944067000000910
是参数化的非线性函数，ηk是过程噪声，μk是测量噪声。
并对两个方程进行初始化，计算状态变量估计以及其协方差：
Figure BDA0001594406700000101
Figure BDA0001594406700000102
其中：
Figure BDA0001594406700000103
是状态值，Pk是协方差矩阵；
步骤S52：引入了一个最小的sigma集合，运用减少Sigma点集方法对所述初始状态变量θk进行Sigma采样，获得n+1个采样点以及权重系数，以减少计算复杂度；随机变量
Figure BDA0001594406700000104
具有均值
Figure BDA0001594406700000105
和协方差矩阵PXX＞0，则：
Figure BDA0001594406700000106
Wweight＝[W ωn+1] (9)
其中：
Figure BDA0001594406700000107
Figure BDA0001594406700000108
步骤S53：状态更新，通过离散时间非线性系统的状态方程将每个采样点的k时刻的最优状态变量的状态估计变换为k+1时刻的状态变量的状态估计
Figure BDA0001594406700000109
并通过合并k+1时刻的状态估计
Figure BDA00015944067000001010
的向量，获得k+1时刻的状态变量的状态先验估计
Figure BDA00015944067000001011
和协方差
Figure BDA00015944067000001012
其中，所述状态估计
Figure BDA00015944067000001013
为：
Figure BDA00015944067000001014
其中，β为放缩因子；
所述状态先验估计
Figure BDA00015944067000001015
为：
Figure BDA0001594406700000111
所述状态变量的协方差
Figure BDA0001594406700000112
为：
Figure BDA0001594406700000113
步骤S54：量测更新，通过离散时间非线性系统的量测方程建立k时刻的状态变量的状态估计
Figure BDA0001594406700000114
和k时刻的量测预测估计
Figure BDA0001594406700000115
之间的联系以完成量测预测，并估计k时刻的量测预测的协方差
Figure BDA0001594406700000116
以及k时刻的状态变量和量测预测之间的协方差
Figure BDA0001594406700000117
其中，所述量测估计
Figure BDA0001594406700000118
为：
Figure BDA0001594406700000119
所述k时刻的量测预测的均值
Figure BDA00015944067000001110
为：
Figure BDA00015944067000001111
所述k时刻的量测预测的协方差
Figure BDA00015944067000001112
为：
Figure BDA00015944067000001113
所述k时刻的状态变量和量测预测之间的协方差
Figure BDA00015944067000001114
为：
Figure BDA00015944067000001115
步骤S55：通过建立协方差
Figure BDA00015944067000001116
和协方差
Figure BDA00015944067000001117
的关系，更新k时刻的状态变量的状态估计和协方差；
所述协方差之间的关系是：
Figure BDA00015944067000001118
通过上述关系对k+1时刻的状态变量的状态估计和协方差进行修正：
Figure BDA0001594406700000121
Figure BDA0001594406700000122
步骤S56：将获得的修正后k+1时刻的状态变量
Figure BDA0001594406700000123
重组BP神经网络模型，并计算此时的模型预测输出与实际输出之间的误差，如果小于既设精度要求，则输出所述神经网络模型的最优状态变量
Figure BDA0001594406700000124
反之，重新进入步骤1
步骤S6：将所述最优状态变量作为所述神经网络模型的wij、vjd、
Figure BDA0001594406700000125
和
Figure BDA0001594406700000126
对式(1)与(2)进行更新，获得所述训练样本更新后的神经网络模型；
步骤S7：将所述测试样本中的
Figure BDA0001594406700000127
输入到更新后的神经网络模型，得到预测结果，将所述预测结果与所述测试样本中的实际输出
Figure BDA0001594406700000128
进行比较，如果比较结果小于预设误差值，所构建的神经网络模型有效；否则重复上述步骤S1-S7，直至所述比较结果小于所述预设误差值为止。
本发明通过几组测试得到如下的技术效果：
图1a-图1b为比较预测结果和运行数据图，其中，图1a在训练阶段的H2S浓度，显示H2S浓度下降的散点图和列车数据集的三种模型估计，图1b在训练阶段的CO2浓度，在预测二氧化碳浓度方面对三种模型的性能进行了比较。
图2a-图2b为对模型精度的比较图，其中，图2a在训练阶段的H2S浓度，图2b训练阶段的CO2浓度。
图3a-图3b为比较预测结果和运行数据图，其中，图3a在训练阶段的H2S浓度，图3b训练阶段的CO2浓度。
对于超过80％的数据点，ScMiUKFNN模型估计的H2S浓度和CO2浓度的绝对相对误差小于10％，验证了所提出的模型的准确性，故所建模有效。
以上所述，仅为本发明的具体实施方式，但本发明的保护范围并不局限于此，任何熟悉本技术领域的技术人员在本发明揭露的技术范围内，可轻易想到变化或替换，都应涵盖在本发明的保护范围之内。因此，本发明的保护范围应所述以权利要求的保护范围为准。

Claims (3)

1.一种基于ScMiUKFNN算法的天然气净化工艺建模方法，其特征在于，本方法包括以下步骤：
步骤S1：选择影响脱硫效率的工艺参数和脱硫单元的性能指标；
步骤S2：采集预设时间的所述工艺参数和所述性能指标的数据，剔除误差样本后形成样本集[X，Y]；
步骤S3：对样本集[X，Y]进行归一化，形成归一化样本集
Figure FDA0003080571930000011
取所述归一化样本集
Figure FDA0003080571930000012
中一部分样本作为训练样本，剩余部分的样本作为测试样本；
步骤S4：基于所述训练样本构建神经网络模型和所述神经网络模型的初始状态变量θk，以及，将所述训练样本中的
Figure FDA0003080571930000013
作为所述神经网络模型的输入，将所述训练样本中的
Figure FDA0003080571930000014
作为所述神经网络模型的输出；
所述神经网络模型为：
Figure FDA0003080571930000015
Figure FDA0003080571930000016
其中，
Figure FDA0003080571930000017
为所述训练样本的矢量样本值，并作为所述神经网络模型的输入；zj作为所述神经网络模型的隐含层输出；
Figure FDA0003080571930000018
作为所述神经网络模型的输出层输出；wij为神经网络模型的输入层到隐含层的神经元的连接权值；
Figure FDA0003080571930000019
为神经网络模型的输入层到隐含层的神经元的阈值；vjd为所述神经网络模型的隐含层到输出层的神经元的连接权值，
Figure FDA00030805719300000110
为所述神经网络模型的隐含层到输出层的神经元的阈值，i＝1，2，…，m；m为神经网络模型的输入层的神经元的数量，s为神经网络模型的隐含层的神经元的数量，h为神经网络模型的输出层的神经元的数量；
应用于神经网络模型各层神经元的非线性激活函数为：
Figure FDA0003080571930000021
fo(x)＝x (4)
所述初始状态变量为：
Figure FDA0003080571930000022
步骤S5：利用ScMiUKFNN算法估计所述神经网络模型的最优状态变量；
所述步骤S5包括：
步骤S51：在建立的所述神经网络模型中，将神经网络模型的权值和阈值组成的状态向量视为算法所需的状态方程，神经网络的模型视为量测方程：
Figure FDA0003080571930000023
其中，
Figure FDA0003080571930000024
为神经网络模型的输入，
Figure FDA0003080571930000025
为神经网络模型的输出，
Figure FDA0003080571930000026
是参数化的非线性函数，ηk是过程噪声，μk是测量噪声；
对状态方程和量测方程进行初始化，计算状态变量估计以及其协方差：
Figure FDA0003080571930000027
Figure FDA0003080571930000028
其中：
Figure FDA0003080571930000029
是状态值，Pk是协方差矩阵；
步骤S52：引入一个最小的sigma集合，运用减少Sigma点集方法对所述初始状态变量θk进行Sigma采样，获得n+1个采样点以及权重系数，以减少计算复杂度；随机变量
Figure FDA00030805719300000210
具有均值
Figure FDA00030805719300000211
和协方差矩阵PXX＞0，则：
Figure FDA00030805719300000212
Wweight＝[W ωn+1] (9)
其中：
Figure FDA0003080571930000031
Figure FDA0003080571930000032
步骤S53：状态更新，通过离散时间非线性系统的状态方程将每个采样点的k时刻的最优状态变量的状态估计变换为k+1时刻的状态变量的状态估计
Figure FDA0003080571930000033
并通过合并k+1时刻的状态估计
Figure FDA0003080571930000034
的向量，获得k+1时刻的状态变量的状态先验估计
Figure FDA0003080571930000035
和协方差
Figure FDA0003080571930000036
其中，所述状态估计
Figure FDA0003080571930000037
为：
Figure FDA0003080571930000038
其中，β为放缩因子，f为线性方程；
所述状态先验估计
Figure FDA0003080571930000039
为：
Figure FDA00030805719300000310
所述状态变量的协方差
Figure FDA00030805719300000311
为：
Figure FDA00030805719300000312
步骤S54：量测更新，通过离散时间非线性系统的量测方程建立k时刻的状态变量的状态估计
Figure FDA00030805719300000313
和k时刻的量测预测估计
Figure FDA00030805719300000314
之间的联系以完成量测预测，并估计k时刻的量测预测的协方差
Figure FDA00030805719300000315
以及k时刻的状态变量和量测预测之间的协方差
Figure FDA00030805719300000316
其中，所述量测估计
Figure FDA00030805719300000317
为：
Figure FDA00030805719300000318
所述k时刻的量测预测的均值
Figure FDA00030805719300000319
为：
Figure FDA0003080571930000041
所述k时刻的量测预测的协方差
Figure FDA0003080571930000042
为：
Figure FDA0003080571930000043
所述k时刻的状态变量和量测预测之间的协方差
Figure FDA0003080571930000044
为：
Figure FDA0003080571930000045
步骤S55：通过建立协方差
Figure FDA0003080571930000046
和协方差
Figure FDA0003080571930000047
的关系，更新k时刻的状态变量的状态估计和协方差；
所述协方差之间的关系是：
Figure FDA0003080571930000048
通过上述关系对k+1时刻的状态变量的状态估计和协方差进行修正：
Figure FDA0003080571930000049
Figure FDA00030805719300000410
步骤S56：将获得的修正后k+1时刻的状态变量
Figure FDA00030805719300000411
重组神经网络模型，并计算此时神经网络模型的预测输出与实际输出之间的误差，如果小于既设精度要求，则输出所述神经网络模型的最优状态变量
Figure FDA00030805719300000412
反之，重新进入步骤1；
步骤S6：将所述最优状态变量作为所述神经网络模型的wij、vjd、
Figure FDA00030805719300000413
和
Figure FDA00030805719300000414
对公式(1)和公式(2)进行更新，获得训练样本更新后的神经网络模型；
步骤S7：将所述测试样本中的
Figure FDA00030805719300000415
输入到更新后的神经网络模型，得到预测结果，将所述预测结果与所述测试样本中的实际输出
Figure FDA00030805719300000416
进行比较，如果比较结果小于预设误差值，所构建的神经网络模型有效；否则重复上述步骤S1-S7，直至所述比较结果小于所述预设误差值为止。
2.根据权利要求1所述的一种基于ScMiUKFNN算法的天然气净化工艺建模方法，其特征在于，所述工艺参数包括进入尾气吸收塔的贫胺液流量、进入二级吸收塔的贫胺液流量、原料气处理量、尾气单元返回脱硫单元的半富胺液流量、一级吸收塔胺液入塔温度、二级吸收塔胺液入塔温度、闪蒸罐压力、一个重沸器的蒸汽消耗量、另一个重沸器的蒸汽消耗量和蒸汽预热器的蒸汽消耗量；脱硫单元的性能指标包括净化气中H2S和CO2的浓度。
3.根据权利要求1所述的一种基于ScMiUKFNN算法的天然气净化工艺建模方法，其特征在于，步骤S3中，取所述归一化样本集
Figure FDA0003080571930000051
中前80％的样本作为训练样本，而剩余的20％样本作为测试样本。
CN201810200655.1A 2018-03-12 2018-03-12 一种基于ScMiUKFNN算法的天然气净化工艺建模方法 Active CN108182337B (zh)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201810200655.1A CN108182337B (zh) 2018-03-12 2018-03-12 一种基于ScMiUKFNN算法的天然气净化工艺建模方法

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201810200655.1A CN108182337B (zh) 2018-03-12 2018-03-12 一种基于ScMiUKFNN算法的天然气净化工艺建模方法

Publications (2)

Publication Number Publication Date
CN108182337A CN108182337A (zh) 2018-06-19
CN108182337B true CN108182337B (zh) 2021-07-09

Family

ID=62553387

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201810200655.1A Active CN108182337B (zh) 2018-03-12 2018-03-12 一种基于ScMiUKFNN算法的天然气净化工艺建模方法

Country Status (1)

Country Link
CN (1) CN108182337B (zh)

Families Citing this family (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN116127729B (zh) * 2022-12-28 2023-08-15 青芥一合碳汇(武汉)科技有限公司 基于线性动态模型的二氧化碳捕集精准预测方法及系统

Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN103345559A (zh) * 2013-07-10 2013-10-09 重庆科技学院 铝电解过程电解槽工艺能耗的动态演化建模方法
CN106777468A (zh) * 2016-11-14 2017-05-31 重庆科技学院 高含硫天然气脱硫工艺强跟踪演化建模方法
CN106777866A (zh) * 2016-11-14 2017-05-31 重庆科技学院 面向节能降耗的高含硫天然气净化工艺建模与优化方法

Family Cites Families (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US7130446B2 (en) * 2001-12-03 2006-10-31 Microsoft Corporation Automatic detection and tracking of multiple individuals using multiple cues

Patent Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN103345559A (zh) * 2013-07-10 2013-10-09 重庆科技学院 铝电解过程电解槽工艺能耗的动态演化建模方法
CN106777468A (zh) * 2016-11-14 2017-05-31 重庆科技学院 高含硫天然气脱硫工艺强跟踪演化建模方法
CN106777866A (zh) * 2016-11-14 2017-05-31 重庆科技学院 面向节能降耗的高含硫天然气净化工艺建模与优化方法

Non-Patent Citations (3)

* Cited by examiner, † Cited by third party
Title
A new smallest sigma set for the Unscented Transform and its applications on SLAM;Henrique M. Menegaz,et al;《IEEE》;20120301;正文第3172-3175页 *
Scaled Minimum Unscented Multiple Hypotheses Mixing Filter;Menegaz H M,et al;《IEEE》;20130816;正文第2460-2465页 *
基于新型卡尔曼滤波的异步电机无传感器控制系统研究;杜清潭;《中国优秀硕士学位论文全文数据库 工程科技II辑》;20150215;正文第4章 *

Also Published As

Publication number Publication date
CN108182337A (zh) 2018-06-19

Similar Documents

Publication Publication Date Title
CN106777866B (zh) 面向节能降耗的高含硫天然气净化工艺建模与优化方法
CN103345559B (zh) 铝电解过程电解槽工艺能耗的动态演化建模方法
CN111191191B (zh) 一种精准预测混凝土坝变形效应的组合模型的构建方法
CN106777465B (zh) 高含硫天然气净化工艺动态演化建模与节能优化方法
CN104636600B (zh) 基于极限学习机的高含硫天然气净化工艺建模、优化方法
CN109670625B (zh) 基于无迹卡尔曼滤波最小二乘支持向量机的NOx排放浓度预测方法
CN105784556B (zh) 一种基于自组织模糊神经网络的空气细颗粒物pm2.5软测量方法
CN105981025A (zh) 基于序列递归滤波三维变分的实测海洋环境数据同化方法
CN106156434B (zh) 基于局部时滞重构的滑动窗时间差-高斯过程回归建模方法
CN107292446B (zh) 一种基于考虑分量关联性小波分解的混合风速预测方法
CN110826791A (zh) 一种基于长短时记忆神经网络的混合风功率预测方法
CN108509692B (zh) 一种基于MiUKFNN算法的高含硫天然气脱硫工艺建模方法
CN115034129B (zh) 一种面向火电厂脱硝装置的NOx排放浓度软测量方法
CN106777466B (zh) 基于st-upfnn算法的高含硫天然气净化工艺的动态演化建模方法
CN108595892A (zh) 基于时间差分模型的软测量建模方法
CN109599866B (zh) 一种预测辅助的电力系统状态估计方法
CN117711523A (zh) 确定气体排放浓度的方法及装置
CN108182337B (zh) 一种基于ScMiUKFNN算法的天然气净化工艺建模方法
CN112379274A (zh) 一种动力电池剩余寿命预测方法
CN109033524B (zh) 一种基于鲁棒混合模型的化工过程浓度变量在线估计方法
CN106777468A (zh) 高含硫天然气脱硫工艺强跟踪演化建模方法
CN112862063A (zh) 一种基于深度信念网络的复杂管网泄漏定位方法
CN110909492A (zh) 一种基于极端梯度提升算法的污水处理过程软测量方法
CN110837886A (zh) 一种基于elm-sl0神经网络的出水nh4-n软测量方法
CN113686810B (zh) 一种基于卷积神经网络的近红外光谱波长选择方法

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
GR01 Patent grant
GR01 Patent grant
EE01 Entry into force of recordation of patent licensing contract

Application publication date: 20180619

Assignee: Guangzhou Senyu automation machinery design Co.,Ltd.

Assignor: Chongqing University of Science & Technology

Contract record no.: X2023980040566

Denomination of invention: A Modeling Method for Natural Gas Purification Process Based on ScMiUKFNN Algorithm

Granted publication date: 20210709

License type: Common License

Record date: 20230830

EE01 Entry into force of recordation of patent licensing contract
EE01 Entry into force of recordation of patent licensing contract

Application publication date: 20180619

Assignee: Foshan shangxiaoyun Technology Co.,Ltd.

Assignor: Chongqing University of Science & Technology

Contract record no.: X2023980041008

Denomination of invention: A Modeling Method for Natural Gas Purification Process Based on ScMiUKFNN Algorithm

Granted publication date: 20210709

License type: Common License

Record date: 20230906

Application publication date: 20180619

Assignee: FOSHAN YAOYE TECHNOLOGY Co.,Ltd.

Assignor: Chongqing University of Science & Technology

Contract record no.: X2023980041004

Denomination of invention: A Modeling Method for Natural Gas Purification Process Based on ScMiUKFNN Algorithm

Granted publication date: 20210709

License type: Common License

Record date: 20230906

Application publication date: 20180619

Assignee: FOSHAN YIQING TECHNOLOGY Co.,Ltd.

Assignor: Chongqing University of Science & Technology

Contract record no.: X2023980041003

Denomination of invention: A Modeling Method for Natural Gas Purification Process Based on ScMiUKFNN Algorithm

Granted publication date: 20210709

License type: Common License

Record date: 20230906

Application publication date: 20180619

Assignee: Guangzhou Fuke Machinery Trade Co.,Ltd.

Assignor: Chongqing University of Science & Technology

Contract record no.: X2023980040999

Denomination of invention: A Modeling Method for Natural Gas Purification Process Based on ScMiUKFNN Algorithm

Granted publication date: 20210709

License type: Common License

Record date: 20230906

Application publication date: 20180619

Assignee: Guangzhou trump Environmental Protection Technology Co.,Ltd.

Assignor: Chongqing University of Science & Technology

Contract record no.: X2023980040995

Denomination of invention: A Modeling Method for Natural Gas Purification Process Based on ScMiUKFNN Algorithm

Granted publication date: 20210709

License type: Common License

Record date: 20230906

EE01 Entry into force of recordation of patent licensing contract
EE01 Entry into force of recordation of patent licensing contract

Application publication date: 20180619

Assignee: Wokang (Guangzhou) Electronic Technology Co.,Ltd.

Assignor: Chongqing University of Science & Technology

Contract record no.: X2023980041471

Denomination of invention: A Modeling Method for Natural Gas Purification Process Based on ScMiUKFNN Algorithm

Granted publication date: 20210709

License type: Common License

Record date: 20230913

Application publication date: 20180619

Assignee: Changhong (Guangzhou) Information Technology Co.,Ltd.

Assignor: Chongqing University of Science & Technology

Contract record no.: X2023980041467

Denomination of invention: A Modeling Method for Natural Gas Purification Process Based on ScMiUKFNN Algorithm

Granted publication date: 20210709

License type: Common License

Record date: 20230913

Application publication date: 20180619

Assignee: Dongguan Yaluo Environmental Protection Technology Co.,Ltd.

Assignor: Chongqing University of Science & Technology

Contract record no.: X2023980041460

Denomination of invention: A Modeling Method for Natural Gas Purification Process Based on ScMiUKFNN Algorithm

Granted publication date: 20210709

License type: Common License

Record date: 20230913

EE01 Entry into force of recordation of patent licensing contract
EE01 Entry into force of recordation of patent licensing contract

Application publication date: 20180619

Assignee: Laishi (Guangzhou) Digital Technology Co.,Ltd.

Assignor: Chongqing University of Science & Technology

Contract record no.: X2023980041991

Denomination of invention: A Modeling Method for Natural Gas Purification Process Based on ScMiUKFNN Algorithm

Granted publication date: 20210709

License type: Common License

Record date: 20230922

Application publication date: 20180619

Assignee: Guangzhou Qiming Technology Co.,Ltd.

Assignor: Chongqing University of Science & Technology

Contract record no.: X2023980041990

Denomination of invention: A Modeling Method for Natural Gas Purification Process Based on ScMiUKFNN Algorithm

Granted publication date: 20210709

License type: Common License

Record date: 20230922

Application publication date: 20180619

Assignee: Guangzhou Daguan Digital Technology Co.,Ltd.

Assignor: Chongqing University of Science & Technology

Contract record no.: X2023980041989

Denomination of invention: A Modeling Method for Natural Gas Purification Process Based on ScMiUKFNN Algorithm

Granted publication date: 20210709

License type: Common License

Record date: 20230922

Application publication date: 20180619

Assignee: Yichang Dae Urban and Rural Construction Co.,Ltd.

Assignor: Chongqing University of Science & Technology

Contract record no.: X2023980041988

Denomination of invention: A Modeling Method for Natural Gas Purification Process Based on ScMiUKFNN Algorithm

Granted publication date: 20210709

License type: Common License

Record date: 20230922

Application publication date: 20180619

Assignee: Guangzhou Dongtong Technology Co.,Ltd.

Assignor: Chongqing University of Science & Technology

Contract record no.: X2023980041866

Denomination of invention: A Modeling Method for Natural Gas Purification Process Based on ScMiUKFNN Algorithm

Granted publication date: 20210709

License type: Common License

Record date: 20230922

Application publication date: 20180619

Assignee: Dongguan Zhaoyi Information Technology Co.,Ltd.

Assignor: Chongqing University of Science & Technology

Contract record no.: X2023980041863

Denomination of invention: A Modeling Method for Natural Gas Purification Process Based on ScMiUKFNN Algorithm

Granted publication date: 20210709

License type: Common License

Record date: 20230922

Application publication date: 20180619

Assignee: Leta (Guangzhou) Technology Co.,Ltd.

Assignor: Chongqing University of Science & Technology

Contract record no.: X2023980041859

Denomination of invention: A Modeling Method for Natural Gas Purification Process Based on ScMiUKFNN Algorithm

Granted publication date: 20210709

License type: Common License

Record date: 20230922

EE01 Entry into force of recordation of patent licensing contract
EE01 Entry into force of recordation of patent licensing contract
EE01 Entry into force of recordation of patent licensing contract

Application publication date: 20180619

Assignee: GUANGZHOU KUAIZHOU INTELLIGENT ENVIRONMENTAL TECHNOLOGY CO.,LTD.

Assignor: Chongqing University of Science & Technology

Contract record no.: X2023980044603

Denomination of invention: A Modeling Method for Natural Gas Purification Process Based on ScMiUKFNN Algorithm

Granted publication date: 20210709

License type: Common License

Record date: 20231031

Application publication date: 20180619

Assignee: Guangzhou Tuyu Technology Co.,Ltd.

Assignor: Chongqing University of Science & Technology

Contract record no.: X2023980044600

Denomination of invention: A Modeling Method for Natural Gas Purification Process Based on ScMiUKFNN Algorithm

Granted publication date: 20210709

License type: Common License

Record date: 20231031

Application publication date: 20180619

Assignee: GUANGZHOU SHANGCHENG TECHNOLOGY Co.,Ltd.

Assignor: Chongqing University of Science & Technology

Contract record no.: X2023980044597

Denomination of invention: A Modeling Method for Natural Gas Purification Process Based on ScMiUKFNN Algorithm

Granted publication date: 20210709

License type: Common License

Record date: 20231031

Application publication date: 20180619

Assignee: GUANGZHOU JUFENG TECHNOLOGY Co.,Ltd.

Assignor: Chongqing University of Science & Technology

Contract record no.: X2023980044596

Denomination of invention: A Modeling Method for Natural Gas Purification Process Based on ScMiUKFNN Algorithm

Granted publication date: 20210709

License type: Common License

Record date: 20231031

Application publication date: 20180619

Assignee: GUANGZHOU XINGYIN TECHNOLOGY Co.,Ltd.

Assignor: Chongqing University of Science & Technology

Contract record no.: X2023980044593

Denomination of invention: A Modeling Method for Natural Gas Purification Process Based on ScMiUKFNN Algorithm

Granted publication date: 20210709

License type: Common License

Record date: 20231031

Application publication date: 20180619

Assignee: GUANGZHOU LVNENG INTELLIGENT TECHNOLOGY Co.,Ltd.

Assignor: Chongqing University of Science & Technology

Contract record no.: X2023980044591

Denomination of invention: A Modeling Method for Natural Gas Purification Process Based on ScMiUKFNN Algorithm

Granted publication date: 20210709

License type: Common License

Record date: 20231031

Application publication date: 20180619

Assignee: Guangzhou Xiaoqing Technology Co.,Ltd.

Assignor: Chongqing University of Science & Technology

Contract record no.: X2023980044587

Denomination of invention: A Modeling Method for Natural Gas Purification Process Based on ScMiUKFNN Algorithm

Granted publication date: 20210709

License type: Common License

Record date: 20231031

Application publication date: 20180619

Assignee: Guangzhou Fangshao Technology Co.,Ltd.

Assignor: Chongqing University of Science & Technology

Contract record no.: X2023980044586

Denomination of invention: A Modeling Method for Natural Gas Purification Process Based on ScMiUKFNN Algorithm

Granted publication date: 20210709

License type: Common License

Record date: 20231031

Application publication date: 20180619

Assignee: Guangzhou star automation equipment Co.,Ltd.

Assignor: Chongqing University of Science & Technology

Contract record no.: X2023980044559

Denomination of invention: A Modeling Method for Natural Gas Purification Process Based on ScMiUKFNN Algorithm

Granted publication date: 20210709

License type: Common License

Record date: 20231031

EE01 Entry into force of recordation of patent licensing contract

Application publication date: 20180619

Assignee: Guangzhou Yuming Technology Co.,Ltd.

Assignor: Chongqing University of Science & Technology

Contract record no.: X2023980047712

Denomination of invention: A modeling method for natural gas purification process based on ScMiUKFNN algorithm

Granted publication date: 20210709

License type: Common License

Record date: 20231124

Application publication date: 20180619

Assignee: Yajia (Guangzhou) Electronic Technology Co.,Ltd.

Assignor: Chongqing University of Science & Technology

Contract record no.: X2023980047706

Denomination of invention: A modeling method for natural gas purification process based on ScMiUKFNN algorithm

Granted publication date: 20210709

License type: Common License

Record date: 20231124

Application publication date: 20180619

Assignee: Guangzhou Yibo Yuntian Information Technology Co.,Ltd.

Assignor: Chongqing University of Science & Technology

Contract record no.: X2023980047705

Denomination of invention: A modeling method for natural gas purification process based on ScMiUKFNN algorithm

Granted publication date: 20210709

License type: Common License

Record date: 20231124

Application publication date: 20180619

Assignee: GUANGZHOU XIAONAN TECHNOLOGY Co.,Ltd.

Assignor: Chongqing University of Science & Technology

Contract record no.: X2023980047703

Denomination of invention: A modeling method for natural gas purification process based on ScMiUKFNN algorithm

Granted publication date: 20210709

License type: Common License

Record date: 20231124

Application publication date: 20180619

Assignee: GUANGZHOU YIDE INTELLIGENT TECHNOLOGY Co.,Ltd.

Assignor: Chongqing University of Science & Technology

Contract record no.: X2023980047702

Denomination of invention: A modeling method for natural gas purification process based on ScMiUKFNN algorithm

Granted publication date: 20210709

License type: Common License

Record date: 20231124

Application publication date: 20180619

Assignee: Lingteng (Guangzhou) Electronic Technology Co.,Ltd.

Assignor: Chongqing University of Science & Technology

Contract record no.: X2023980047701

Denomination of invention: A modeling method for natural gas purification process based on ScMiUKFNN algorithm

Granted publication date: 20210709

License type: Common License

Record date: 20231124

Application publication date: 20180619

Assignee: Guangzhou Taipu Intelligent Technology Co.,Ltd.

Assignor: Chongqing University of Science & Technology

Contract record no.: X2023980047700

Denomination of invention: A modeling method for natural gas purification process based on ScMiUKFNN algorithm

Granted publication date: 20210709

License type: Common License

Record date: 20231124

Application publication date: 20180619

Assignee: Yuxin (Guangzhou) Electronic Technology Co.,Ltd.

Assignor: Chongqing University of Science & Technology

Contract record no.: X2023980047695

Denomination of invention: A modeling method for natural gas purification process based on ScMiUKFNN algorithm

Granted publication date: 20210709

License type: Common License

Record date: 20231124

EE01 Entry into force of recordation of patent licensing contract
EE01 Entry into force of recordation of patent licensing contract

Application publication date: 20180619

Assignee: Guangxi GaoMin Technology Development Co.,Ltd.

Assignor: Chongqing University of Science & Technology

Contract record no.: X2023980053986

Denomination of invention: A modeling method for natural gas purification process based on ScMiUKFNN algorithm

Granted publication date: 20210709

License type: Common License

Record date: 20231227

EE01 Entry into force of recordation of patent licensing contract
EE01 Entry into force of recordation of patent licensing contract

Application publication date: 20180619

Assignee: Yuao Holdings Co.,Ltd.

Assignor: Chongqing University of Science & Technology

Contract record no.: X2024980000640

Denomination of invention: A modeling method for natural gas purification process based on ScMiUKFNN algorithm

Granted publication date: 20210709

License type: Common License

Record date: 20240119

EE01 Entry into force of recordation of patent licensing contract
EE01 Entry into force of recordation of patent licensing contract

Application publication date: 20180619

Assignee: Foshan chopsticks Technology Co.,Ltd.

Assignor: Chongqing University of Science & Technology

Contract record no.: X2024980003017

Denomination of invention: A modeling method for natural gas purification process based on ScMiUKFNN algorithm

Granted publication date: 20210709

License type: Common License

Record date: 20240322

Application publication date: 20180619

Assignee: Foshan qianshun Technology Co.,Ltd.

Assignor: Chongqing University of Science & Technology

Contract record no.: X2024980003012

Denomination of invention: A modeling method for natural gas purification process based on ScMiUKFNN algorithm

Granted publication date: 20210709

License type: Common License

Record date: 20240322

EE01 Entry into force of recordation of patent licensing contract
EE01 Entry into force of recordation of patent licensing contract

Application publication date: 20180619

Assignee: Foshan helixing Technology Co.,Ltd.

Assignor: Chongqing University of Science & Technology

Contract record no.: X2024980004524

Denomination of invention: A modeling method for natural gas purification process based on ScMiUKFNN algorithm

Granted publication date: 20210709

License type: Common License

Record date: 20240419

EE01 Entry into force of recordation of patent licensing contract
EE01 Entry into force of recordation of patent licensing contract
EE01 Entry into force of recordation of patent licensing contract

Application publication date: 20180619

Assignee: Qishi (Yantai) data Technology Co.,Ltd.

Assignor: Chongqing University of Science & Technology

Contract record no.: X2024980008104

Denomination of invention: A modeling method for natural gas purification process based on ScMiUKFNN algorithm

Granted publication date: 20210709

License type: Common License

Record date: 20240701

Application publication date: 20180619

Assignee: Yantai Xingyue coating equipment Co.,Ltd.

Assignor: Chongqing University of Science & Technology

Contract record no.: X2024980008099

Denomination of invention: A modeling method for natural gas purification process based on ScMiUKFNN algorithm

Granted publication date: 20210709

License type: Common License

Record date: 20240701

Application publication date: 20180619

Assignee: Yantai Zhonglan Environmental Technology Co.,Ltd.

Assignor: Chongqing University of Science & Technology

Contract record no.: X2024980008084

Denomination of invention: A modeling method for natural gas purification process based on ScMiUKFNN algorithm

Granted publication date: 20210709

License type: Common License

Record date: 20240701