CN105301985A - Method and system for measuring length of calcium carbide furnace electrode - Google Patents

Method and system for measuring length of calcium carbide furnace electrode Download PDF

Info

Publication number
CN105301985A
CN105301985A CN 201510819613 CN201510819613A CN105301985A CN 105301985 A CN105301985 A CN 105301985A CN 201510819613 CN201510819613 CN 201510819613 CN 201510819613 A CN201510819613 A CN 201510819613A CN 105301985 A CN105301985 A CN 105301985A
Authority
CN
Grant status
Application
Patent type
Prior art keywords
electrode
consumption
method
calcium
carbide
Prior art date
Application number
CN 201510819613
Other languages
Chinese (zh)
Inventor
苏宏业
张树吉
古勇
金晓明
张立
何忠
周军
Original Assignee
浙江中控软件技术有限公司
新疆天业(集团)有限公司
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

Links

Abstract

The invention provides a method for detecting the length of a calcium carbide furnace electrode. The method comprises the steps: obtaining a non-linear prediction model of electrode consumption according to a model training sample set, wherein electrode temperature T, electrode power P, a furnace charge ratio R and calcium carbide output G are sample input variables, and the electrode consumption is a sample output variable; determining electrode consumption in a current working condition according to the non-linear prediction model of electrode consumption; and determining the length of an electrode according to the electrode consumption in the current working condition. The method can monitor the position changes of an electrode in real time and then guide pressing and releasing for electrodes, so the pressing and releasing amount for electrodes can be accurately controlled, and electrodes are made to work in a balance manner. The method guarantees stable production and energy-saving operation of a calcium carbide furnace.

Description

一种电石炉电极长度的测量方法及系统 Method and system for measuring a calcium carbide furnace electrode length

技术领域 FIELD

[0001 ] 本发明涉及工业控制信息领域,特别涉及一种电石炉电极长度的测量方法及系统。 [0001] The present invention relates to the field of industrial control information, in particular, relates to a method and a system for measuring the length of the electrode carbide furnace.

背景技术 Background technique

[0002] 电石炉是生产电石的主体设备,其主要是通过电极的埋弧电热和物料的电阻电热来加热炉内物料,如生石灰和碳素等,使物料反应生成电石。 [0002] The production of calcium carbide is calcium carbide furnace apparatus main body, which is primarily to the heating furnace material, such as lime and the like carbon submerged electric resistance heating electrodes and the material, the material reacts carbide.

[0003] 电石炉中常采用的电极包括自焙电极、石墨电极或碳素电极,电极把大电流输送至炉内,在电极的末端产生电弧,进而将电能转换为热能,由于电弧上产生了极高的温度, 电极端头升华,导致电极不断消耗而变短,因此,需要定期压放电极以保持一定的电极长度,使其能够稳定做功,保证产出电石的质量。 An electrode [0003] carbide furnace often used include self-baking electrode, a graphite electrode or carbon electrode, the electrode is fed to the furnace large current, an arc is generated at the tip of the electrode, in turn converted to heat energy due to the generation of the arc electrode high temperatures, electrode tip sublimation, resulting in consumption of the electrode becomes short continuously, and therefore, the need for regular pressure discharge electrode to maintain a constant length of the electrode, it is possible to stably work to ensure that the quality of output carbide.

[0004] 目前,电极的压放操作,即电极压放间隔时间的设置,主要依赖于生产的经验,人为的因素较大,这会导致各相电极的插入深度存在差异,做功不平衡,继而影响电石质量和电耗。 [0004] Currently, the discharge operation voltage electrode, i.e. an electrode disposed pressure release interval, primarily depends on the experience of the production of larger human factors, which can lead to differences in the electrode insertion depth of each phase, acting imbalance, then calcium carbide affect the quality and power consumption. 若能实时获取电极长度,监控电极的位置变化,进而指导电极的压放操作,以精确的控制各电极的压放量,使得各电极做功平衡,为电石炉的平稳生产和节能运行提供保障。 Acquired in real time if the length of the electrode, monitoring the change in position of the electrodes, and thus the pressure release operation guide electrode, to precisely control the pressure of the volume of each electrode so that each electrode work balance, to provide protection and energy-saving operation for the smooth production of calcium carbide furnace.

[0005] 然而,目前主要是在停炉期间由人工测量获得电极长度,开炉运行期间还无法实时检测到电极长度的变化,因此,有必要提出一种电极长度的检测方法,能够实时获得电极的长度。 [0005] However, the key is obtained by the electrode length measured manually during shutdown, during the opening operation of the burner can not detect in real time a change in length of the electrode, therefore, necessary to propose a method for detecting the length of the electrode, the electrode can be obtained in real time length.

发明内容 SUMMARY

[0006] 有鉴于此,本发明的目的在于提供一种电石炉电极长度的测量方法,实现实时获取电极长度,为电石炉的平稳生产和节能运行提供保障。 [0006] In view of this, an object of the present invention to provide a method for measuring the length of the calcium carbide furnace electrode, the electrode length to achieve real-time access, to provide protection and energy-saving operation for the smooth production of calcium carbide furnace.

[0007] 为实现上述目的,本发明有如下技术方案: [0007] To achieve the above object, the present invention has the following technical solutions:

[0008] -种电石炉电极长度的检测方法,所述方法包括: [0008] - detection electrode length carbide furnaces, the method comprising species:

[0009] 基于模型训练样本集,获取电极消耗量的非线性预测模型,其中,电极温度T、电极功率P、炉料配比R和电石出炉量G为样本输入变量,电极消耗量为样本输出变量; [0009] Based on the model training sample set, acquired linear predictive model consumption electrode, wherein the electrode temperature T, the electrodes of the power P, R and the ratio of the charge amount G baked carbide sample input variables, output variables sample electrode consumption ;

[0010] 根据电极消耗量的非线性预测模型,确定当前工况下的电极消耗量; [0010] The electrode consumption linear predictive model to determine the current condition of electrode consumption;

[0011] 由当前工况下的电极消耗量,确定当前电极长度。 [0011] The current consumption by the operating conditions of the electrode, determining a current length of the electrode.

[0012] 可选的,基于模型训练样本集,获取电极消耗量的非线性预测模型的步骤包括: Step [0012] Alternatively, based on the model training sample set, acquired linear predictive model consumption electrode comprising:

[0013] 通过非线性映射将模型训练样本集的输入样本映射到特征空间,构造该特征空间的最优线性回归函数; [0013] by non-linear model mapping the input training set samples are mapped into feature space, an optimal configuration of the feature space linear regression functions;

[0014] 利用结构风险最小化原则,建立最优线性回归函数的目标优化函数,在目标优化函数中,选择训练误差的2范数作为损失函数,且采用等式约束条件; [0014] With the structural risk minimization principle, the goal of establishing the optimal linear regression function optimization function, optimizing the objective function, select the training error norm as a function of loss, and the use of equality constraints;

[0015] 根据拉格朗日乘子,建立目标优化函数的拉格朗日函数; [0015] According to Lagrange multipliers, the establishment-objective optimization function Lagrangian function;

[0016] 由最小二乘法,确定基于拉格朗日乘子的模型; [0016] by the method of least squares, Lagrange multiplier is determined based on the model;

[0017] 将基于拉格朗日乘子的模型中的函数设定为核函数,从而获得电极消耗量的非线性预测模型。 [0017] The kernel function is set based on a model of the Lagrange multiplier function, thereby obtaining a linear predictive model electrode consumption.

[0018] 可选的,所述核函数为径向基核函数。 [0018] Alternatively, the kernel function is a radial basis kernel function.

[0019] 可选的,所述模型训练样本集为多个,在获得电极消耗量的非线性预测模型之后, 还包括: [0019] Optionally, the model training set into a plurality of, after obtaining the linear predictive model electrode consumption, further comprising:

[0020] 基于误差训练样本集,获得电极消耗量的非线性预测模型的误差评价指标; [0020] Based on the training sample set error, the error evaluation obtained linear predictive model electrode consumption;

[0021] 确定误差评价指标最小的非线性预测模型为电极长度检测中的电极消耗量的非线性预测模型。 [0021] linear predictive model to determine the minimum error for the linear predictive model evaluation index electrode consumption detection electrode length.

[0022] 可选的,所述模型训练样本集经过数据预处理,其中,电极温度T、电极功率P为当前时刻前第一预定时间范围内的平均值,炉料配比R为当前时刻前第二预定时间范围内的平均值,电石出炉量G为每批电石出炉锅数与单锅电石重量的乘积。 [0022] Optionally, the model training set preprocessed data, wherein the electrode temperature T, the electrodes of the power P before the current time point is the average of the first predetermined time range, the ratio of the charge before the current time point is the first R the average value of the second predetermined time range, the amount of released calcium carbide furnace product of G and the number of the pot by weight calcium carbide single pot of each batch of calcium carbide.

[0023] 可选的,当前电极长度11 = !1。 [0023] Alternatively, the current electrode length = 11! 1. +&!^+&(1*(:-&扎,1',其中:!1为当前电极长度,!10为前一时刻电极长度,△扎」为电极升降量,△ d为电极单次压放量,C为压放次数,△ Hxh为电极消耗量,T为电极长度H和H。对应时刻的时间差。 + ^ + (1 * (: -! & Tie, 1 ', wherein: 1 is the current length of the electrode, 10 is the previous time length of the electrode, △ tie "for the electrode lifting amount, △ d electrode single!! pressure volume, C is the number of pressure release, △ Hxh electrode consumption, T is H. H and electrode length corresponding to the time difference between the time.

[0024] 此外,本发明还提供了一种电石炉电极长度的检测系统,包括: [0024] Further, the present invention also provides a furnace electrode length carbide detection system, comprising:

[0025] 电极消耗量预测模型获取模块,用于基于模型训练样本集,获取电极消耗量的非线性预测模型,其中,电极温度T、电极功率P、炉料配比R和电石出炉量G为样本输入变量, 电极消耗量为样本输出变量; [0025] The electrode consumption prediction model acquiring means for training sample set based on the model, obtaining an electrode consumption linear predictive model, wherein the electrode temperature T, the electrodes of the power P, R and the ratio of the charge amount G baked carbide sample input variables, output variables electrode consumption of the sample;

[0026] 电极消耗量确定模块,用于根据电极消耗量的非线性预测模型,确定当前工况下的电极消耗量; [0026] The electrode consumption amount determining module for linear predictive model electrode consumption, determining the current operating conditions of the electrode consumption;

[0027] 当前电极长度确定模块,用于由当前工况下的电极消耗量,确定当前电极长度。 [0027] The current electrode length determining means for consumption by the electrode current operating conditions, determining the current length of the electrode.

[0028] 可选的,电极消耗量预测模型获取模块包括: [0028] Alternatively, the electrode consumption prediction model acquiring module comprises:

[0029] 通过非线性映射将模型训练样本集的输入样本映射到特征空间,构造该特征空间的最优线性回归函数; [0029] by non-linear model mapping the input training set samples are mapped into feature space, an optimal configuration of the feature space linear regression functions;

[0030] 利用结构风险最小化原则,建立最优线性回归函数的目标优化函数,在目标优化函数中,选择训练误差的2范数作为损失函数,且采用等式约束条件; [0030] With the structural risk minimization principle, the goal of establishing the optimal linear regression function optimization function, optimizing the objective function, select the training error norm as a function of loss, and the use of equality constraints;

[0031] 根据拉格朗日乘子,建立目标优化函数的拉格朗日函数; [0031] According to Lagrange multipliers, the establishment-objective optimization function Lagrangian function;

[0032] 由最小二乘法,通过拉格朗日函数确定基于拉格朗日乘子和核函数的电极消耗量的非线性预测模型。 [0032], nonlinear predictive model based on Lagrange multiplier and the electrode consumption is determined by the kernel function by the least square method Lagrange function.

[0033] 可选的,所述核函数为径向基核函数。 [0033] Alternatively, the kernel function is a radial basis kernel function.

[0034] 可选的,所述模型训练样本集为多个,还包括: [0034] Optionally, the model training set into a plurality, further comprising:

[0035] 误差评价模块,基于误差训练样本集,获得电极消耗量的非线性预测模型的误差评价指标;并确定误差评价指标最小的非线性预测模型为电极长度检测中的电极消耗量的非线性预测模型。 [0035] The error estimation module, an error based on the training sample set obtained linear predictive model error evaluation of electrode consumption; determining an error evaluation and linear predictive model is minimal electrode consumption detection electrode length Nonlinear predictive models.

[0036] 可选的,还包括数据预处理模块,用于将模型训练样本集进行数据预处理,其中, 电极温度T、电极功率P为当前时刻前第一预定时间范围内的平均值,炉料配比R为当前时刻前第二预定时间范围内的平均值,电石出炉量G为每批电石出炉锅数与单锅电石重量的乘积。 [0036] Optionally, further comprising a data pre-processing module, model training set for data preprocessing, wherein the electrode temperature T, the electrodes of the power P as mean time before the current within a first predetermined time range, the charge R is a ratio of a second predetermined time before the current average value, G is the product of the amount of released calcium carbide, calcium carbide furnace pot batch number and a single pot by weight of calcium carbide in the time range.

[0037] 可选的,当前电极长度确定模块中,当前电极长度H = !1。 [0037] Alternatively, the electrode length determination module current, the current electrode length H =! 1. +&!^+&(1*(:-&!^*1',其中:H为当前电极长度,H。为前一时刻电极长度,AHsjS电极升降量,Ad为电极单次压放量,C为压放次数,Δ 1为电极消耗量,T为电极长度H和H。对应时刻的时间差。 + ^ + (1 * (: -!! & ^ * 1 ', wherein: H is the current electrode length, H electrode length, AHsjS electrode lifting amount, the Ad electrode single pressure volume as the previous time, C. the number of pressure release, Δ 1 is the consumption of the electrode, T is H. H and electrode length corresponding to the time difference between the time.

[0038] 本发明实施例提供的电石炉电极长度的检测方法及系统,通过电极温度Τ、电极功率Ρ、炉料配比R和电石出炉量G这些样本输入变量,获取电极消耗量的非线性预测模型,通过该非线性预测模型可以获得当前工况下的电极消耗量,继而,由电极消耗量确定出当前电极长度,实现电极长度的软测量。 [0038] The embodiment of the present invention detects linear prediction method and the length of the electrode system provided in the calcium carbide furnace, G these samples Τ input variables by the electrode temperature, electrode power [rho], and calcium carbide furnace charge amount ratio R, acquired electrode consumption model, the current consumption of the electrode can be obtained by the condition linear predictive model, then, the current consumption of the electrodes is determined by the length of the electrode, the length of the soft-measuring electrode. 这样,可以实时监控电极的位置变化,进而指导电极的压放操作,以精确的控制各电极的压放量,使得各电极做功平衡,为电石炉的平稳生产和节能运行提供保障。 Thus, it is possible to monitor in real time the change in position of the electrodes, and thus the pressure release operation guide electrode, to precisely control the pressure of the volume of each electrode so that each electrode work balance, to provide protection and energy-saving operation for the smooth production of calcium carbide furnace.

附图说明 BRIEF DESCRIPTION

[0039] 为了更清楚地说明本发明实施例或现有技术中的技术方案,下面将对实施例或现有技术描述中所需要使用的附图作简单地介绍,显而易见地,下面描述中的附图是本发明的一些实施例,对于本领域普通技术人员来讲,在不付出创造性劳动的前提下,还可以根据这些附图获得其他的附图。 [0039] In order to more clearly illustrate the technical solutions in the embodiments or the prior art embodiment of the present invention, briefly introduced hereinafter, embodiments are described below in the accompanying drawings or described in the prior art needed to be used in describing the embodiments figures some embodiments of the present invention, those of ordinary skill in the art is concerned, without creative efforts, can derive from these drawings other drawings.

[0040] 图1为本发明实施例的电石炉电极长度的检测方法的流程图; [0040] FIG. 1 is a flowchart of the method for detecting electrode length carbide furnace according to an embodiment of the present invention;

[0041] 图2为本发明实施例的检测方法中获取电极消耗量的非线性预测模型的流程图; [0041] FIG 2 is a flowchart of the detection method of linear predictive model acquiring electrode consumption embodiment of the invention;

[0042] 图3为根据本发明实施例的电石炉电极长度的检测系统的结构示意图; [0042] FIG. 3 is a schematic structural diagram of the detection system of the furnace according to the length of the electrode carbide embodiment of the present invention;

[0043] 图4为根据本发明另一实施例的电石炉电极长度的检测系统的结构示意图。 [0043] FIG. 4 is a schematic diagram of the configuration of the detection system of the electrode length carbide furnace according to another embodiment of the present invention.

具体实施方式 detailed description

[0044] 为使本发明实施例的目的、技术方案和优点更加清楚,下面将结合本发明实施例中的附图,对本发明实施例中的技术方案进行清楚、完整地描述,显然,所描述的实施例是本发明一部分实施例,而不是全部的实施例。 [0044] In order that the invention object, technical solutions, and advantages of the embodiments more clearly, the following the present invention in the accompanying drawings, technical solutions of embodiments of the present invention are clearly and completely described, obviously, the described the embodiment is an embodiment of the present invention is a part, but not all embodiments. 基于本发明中的实施例,本领域普通技术人员在没有做出创造性劳动前提下所获得的所有其他实施例,都属于本发明保护的范围。 Based on the embodiments of the present invention, all other embodiments of ordinary skill in the art without any creative effort shall fall within the scope of the present invention.

[0045] 正如背景技术的描述,目前主要是在停炉期间由人工测量获得电极长度,开炉运行期间还无法实时检测到电极长度的变化,电极的压放操作,即电极压放间隔时间的设置, 主要依赖于生产的经验,人为的因素较大,这会导致各相电极的插入深度存在差异,做功不平衡,继而影响电石质量和电耗。 [0045] As described in the background art, the key is obtained during shutdown measuring electrode length manually, during the opening operation of the burner can not detect in real time a change in length of the electrode, the electrode pressure release operation, i.e. pressure discharge electrode interval provided, mainly depends on the experience of the production of larger human factors, which can lead to differences in the electrode insertion depth of each phase, acting imbalance, thereby affecting the quality of calcium carbide and power consumption. 为此,本发明提出了一种电石炉电极长度的检测方法,参考图1所示,所述方法包括: To this end, the present invention provides a method for detecting a calcium carbide furnace electrode length, with reference to Figure 1, the method comprising:

[0046] 基于模型训练样本集,获取电极消耗量的非线性预测模型,其中,电极温度Τ、电极功率P、炉料配比R和电石出炉量G为样本输入变量,电极消耗量为样本输出变量; [0046] Based on the model training sample set, acquired linear predictive model electrode consumption, wherein Τ electrode temperature, electrode power P, R and the ratio of the charge amount G baked carbide sample input variables, output variables sample electrode consumption ;

[0047] 根据电极消耗量的非线性预测模型,确定当前工况下的电极消耗量; [0047] The electrode consumption linear predictive model to determine the current condition of electrode consumption;

[0048] 由当前工况下的电极消耗量,确定当前电极长度。 [0048] The conditions of the current consumption by the electrodes, the current length of the electrode is determined.

[0049] 在该方法中,通过电极温度T、电极功率P、炉料配比R和电石出炉量G这些样本输入变量,获取电极消耗量的非线性预测模型,通过该非线性预测模型可以获得当前工况下的电极消耗量,继而,由电极消耗量确定出当前电极长度,实现电极长度的软测量。 [0049] In this method, the electrode temperature T, the electrodes of the power P, R and carbide furnace charging mixing amount G of these sample input variables, acquired linear predictive model electrode consumption, the current may be obtained by the linear predictive model electrode consumption conditions, which in turn, is determined by the current consumption of the electrode the electrode length, the length of the soft-measuring electrode. 这样,可以实时监控电极的位置变化,进而指导电极的压放操作,以精确的控制各电极的压放量,使得各电极做功平衡,为电石炉的平稳生产和节能运行提供保障。 Thus, it is possible to monitor in real time the change in position of the electrodes, and thus the pressure release operation guide electrode, to precisely control the pressure of the volume of each electrode so that each electrode work balance, to provide protection and energy-saving operation for the smooth production of calcium carbide furnace.

[0050] 为了更好的理解本发明的技术方案和技术效果,以下将结合具体流程图对具体的实施例进行详细的描述。 [0050] In order to better understand the technical solutions and technical effects of the present invention, the following detailed flowchart of binding specific embodiments described in detail.

[0051] 首先,在步骤S101,基于模型训练样本集,获取电极消耗量的非线性预测模型,其中,电极温度T、电极功率P、炉料配比R和电石出炉量G为样本输入变量,电极消耗量为样本输出变量。 [0051] First, at step S101, the model based on the training sample set, acquired linear predictive model consumption electrode, wherein the electrode temperature T, the electrodes of the power P, R and the ratio of the charge amount G baked carbide sample input variables, electrode consumption of sample output variables.

[0052] 在本发明中,选择电极温度T、电极功率P、炉料配比R和电石出炉量G为样本输入变量,和电极消耗量为样本输出变量的模型训练样本集,来确定电极消耗量的非线性预测模型。 [0052] In the present invention, the selective electrode temperature T, the electrodes of the power P, R and the ratio of the charge amount G is baked carbide sample input variables, and the electrode consumption model training sample set as a sample output variables, to determine the consumption of the electrode the non-linear predictive model.

[0053] 其中,电极温度T表征电极软硬程度;电极功率P表征电极做功大小;炉料配比R 为每IOOkg石灰所配用的碳素量,表征炉料中碳素量的大小,电石出炉量G为每批电石的出炉量,表征电石炉实际运行负荷高低。 [0053] wherein, the electrode temperature T soft and hard Characterization electrode; calcium carbide furnace charging mixing amount of the carbon amount per IOOkg R is equipped with the lime, the amount of carbon in the charge characterization size; Characterization electrode work electrode size power P G is released per batch of calcium carbide, calcium carbide furnace characterizing the actual operating load level. 通过这些样本输入的样本集,获取电极消耗量的非线性预测模型。 These sample sets of input samples, obtain linear predictive model electrode consumption.

[0054] 在本发明实施例中,样本集可以包括模型训练样本集和误差训练样本集,还可以进一步包括验证样本集,模型训练样本集可以为多个,为了提高预测模型的预测精度,所取样本数据具有代表性,并尽可能覆盖电石炉生产过程中正常工作范围和不同运行负荷下的运行数据,可以按照如下的结构组成样本,并收集样本数据,样本表达式为Ix 1, yj,样本的数据结构如下表1所示: [0054] In an embodiment of the present invention, the sample set may include a model training set and the error training set may further include a validation sample set, the model may be a plurality of training samples, in order to improve the prediction accuracy of the prediction model, the this representative sample data, and operational data covering as calcium carbide furnace under normal operating range of the production process and the different operating loads, the sample may be composed according to the following structure, and collect sample data, the sample expressed as Ix 1, yj, sample data structure is shown below in table 1:

[0055] 表1样本数据结构 [0055] Table 1 sample data structure

Figure CN105301985AD00071

[0057] 其中,X1为样本的输入,即选取的辅助变量一一电极温度T、电极功率P、炉料配比R、电石出炉量G。 [0057] wherein, the X1 is the input sample, i.e., the selected auxiliary electrode eleven variables temperature T, the electrodes of the power P, the charge ratio R, the amount of released calcium carbide G. 样本的输出yi为待估计的主导变量,即电极消耗量AHxh,电极消耗量的实际值可根据两次停炉期间的电极长度测量值和电极总压放量计算得到。 Output sample yi is the dominant variables to be estimated, i.e., the electrode consumption AHxh, the actual value of the electrode consumption can be calculated according to the electrode and the electrode length measurement volume total pressure during shutdown twice.

[0058] 为了消除数据的噪声,可以对所采集的数据样本进行数据预处理,具体的,电极温度T、电极功率P为当前时刻前第一预定时间范围内的平均值,炉料配比R为当前时刻前第二预定时间范围内的平均值,电石出炉量G为每批电石出炉锅数与单锅电石重量的乘积。 [0058] In order to eliminate the noise data, the data can be pre-collected data sample, specifically, the electrode temperature T, the electrodes of the power P as mean time before the current within a first predetermined range of time, the charge ratio is R a second average value of the current time before the predetermined time range, the amount of released calcium carbide furnace product of G and the number of the pot by weight calcium carbide single pot of each batch of calcium carbide. 预处理之后,可以剔除明显错误和无效的数据,消除噪声,提高所获取的模型的精确性。 After pretreatment, you can eliminate obvious errors and invalid data, eliminating noise, improve the accuracy of the model obtained.

[0059] 在本实施例中,参考图2所示,通过以下具体的步骤获取基于每一个模型训练样本集的电极消耗量的非线性预测模型: [0059] In the present embodiment, with reference to Figure acquired linear predictive model based on the electrode consumption model for each training sample set 2 by the following specific steps:

[0060] 在步骤S201,通过非线性映射将模型训练样本集的输入样本映射到特征空间,构造该特征空间的最优线性回归函数。 [0060] In step S201, the nonlinear mapping by the model training set of input samples is mapped to the feature space, an optimal configuration of the feature space linear regression functions.

[0061] 对于以[电极温度T,电极功率P,炉料配比R,电石出炉量G]T作为模型输入X1,电极消耗量AHxh作为模型输出yi,构成训练模型样本集 [0061] In respect to [the electrode temperature T, the electrodes of the power P, the charge ratio R, the amount of carbide baked G] T as the model inputs X1, electrode consumption as a model AHxh output yi, constituting the training sample set model

Figure CN105301985AD00072

y# R。 y # R. 先通过一非线性映射φ (·)把输入样本从输入空间R 4映射到特征空间 First by a non-linear mapping φ (·) R input samples from the input space is mapped into feature space 4

Figure CN105301985AD00073

在这个高维特征空间中构造最优线性回归函数: Optimal linear regression function configured in this high dimensional feature space:

Figure CN105301985AD00074

[0063] 式中:wT为权向量;b为偏置量。 [0063] wherein: wT is the weight vector; B is the offset amount.

[0064] 而后,在步骤S202,利用结构风险最小化原则,建立最优线性回归函数的目标优化函数,在目标优化函数中,选择训练误差的2范数作为损失函数,且采用等式约束条件。 [0064] Then, at step S202, using the structural risk minimization principle, the goal of establishing the optimal linear regression function optimization function, optimizing the objective function, select the training error norm as a function of loss, and the use of equality constraints . [0065] 在该步骤中,将非线性估计函数转化为高维特征空间中的线性估计问题,利用结构风险最小化原则,寻找w,b就是最小化. [0065] In this step, the non-linear estimation function into a linear estimation problem in the high dimensional feature space using the structural risk minimization principle, looking w, b is minimized.

Figure CN105301985AD00081

控制模型的复杂度,C是正规化参数,1?_为误差控制函数,也即ε不敏感损失函数。 The complexity of the model control, C is the regularization parameter 1? _ Is the error control function, i.e. ε insensitive loss function. 在该优化中损失函数为训练误差的2范数。 In this loss optimization function 2 norm training error. 这样,建立最优线性回归函数的目标优化函数为: Thus, the goal of establishing optimal linear regression function optimization function is:

Figure CN105301985AD00082

[0067] 等式约束条件 [0067] equality constraints

Figure CN105301985AD00083

[0068] 接着,在步骤S203,根据拉格朗日乘子,建立目标优化函数的拉格朗日函数。 [0068] Next, at step S203, the Lagrange multipliers according to the established target Lagrangian optimization function.

[0069] 在该步骤中通过拉格朗日法求解该优化问题,建立拉格朗日函数为: [0069] Lagrangian method for solving this optimization problem established by the Lagrangian function in this step is:

Figure CN105301985AD00084

[0071] 其中ayi,」=1,2,···,1是拉格朗日乘子。 [0071] wherein Ayi, "= 1,2, ..., 1 is the Lagrange multiplier.

[0072] 而后,在步骤S204,由最小二乘法,通过拉格朗日函数确定基于拉格朗日乘子和核函数的电极消耗量的非线性预测模型。 [0072] Then, at step S204, by the least squares method, a linear predictive model to determine the consumption of the electrodes and the Lagrange multiplier kernel based on Lagrange function.

[0073] 根据优化条件: [0073] The optimum conditions:

Figure CN105301985AD00085

[0077] 通过上式,并定义核函数 [0077] By the above formula, and the kernel function is defined

Figure CN105301985AD00086

将优化问题转化为求解线性方程: The optimization problem into solving linear equations:

Figure CN105301985AD00087

[0079] 由最小二乘法获得回归系数Ct1和偏差b,获得基于拉格朗日乘子的模型参数 [0079] Ct1 obtained regression coefficient b by the least squares method and deviations to obtain the model parameters based on Lagrangian multipliers

[bai α 2…a J,进而获得电极消耗量的非线性预测模型: [Bai α 2 ... a J, in turn obtained linear predictive model electrode consumption:

Figure CN105301985AD00088

[0081] 对于核函数K (Xl,Xj),它是满足Metcer条件的任意对称核函数,优选的,所述核函数为径向基核函数,最后,获得电极消耗量的非线性预测模型: [0081] For the kernel function K (Xl, Xj), which satisfies the condition Metcer any symmetric kernel function, preferably, the kernel function is a radial basis kernel function, Finally, a linear predictive model electrode consumption:

Figure CN105301985AD00091

[0083] 对于多组模型训练样本集,可以分别获得基于该组模型训练样本集的电极消耗量的非线性预测模型。 [0083] For the plurality of sets of model training sample set, nonlinear predictive model based on the set of electrode consumption model training set can be obtained, respectively.

[0084] 接着,可以基于误差训练样本集,对上述的非线性预测模型进行误差评价,从而, 从其中选择更优的模型作为电极长度检测中的电极消耗量的非线性预测模型。 [0084] Next, based on the error training set, the above-described linear predictive model for evaluating an error, so, more preferably selected from among linear predictive model as a model of electrode consumption detection electrode length.

[0085] 具体的,首先,基于误差训练样本集,获得电极消耗量的非线性预测模型的误差评价指标。 [0085] Specifically, first, based on the error training set to obtain linear predictive model error evaluation electrode consumption.

[0086] 基于误差训练样本集S,包含样本数据为1,定义误差函数: [0086] Based on the training sample set error S, comprising sample data is 1, the error function is defined:

Figure CN105301985AD00092

[0089] 其中:i = 1,2,…,1。 [0089] where: i = 1,2, ..., 1. 选择最终误差评价函数为: Select the final error evaluation function:

[0090] e ( γ,δ ) = min G1+ η e2) [0090] e (γ, δ) = min G1 + η e2)

[0091] 式中:γ为误差惩罚参数;δ为核参数;η为权重参数。 [0091] where: γ is the error penalty parameter; [delta] nuclear parameter; [eta] is a weight parameter.

[0092] 根据经验选择均方差和最大方差的权重,一般可选择n = 1。 [0092] The variance weights are selected empirically and maximum variance weight, generally choose n = 1. 利用上述误差评价函数获得误差评价指标。 Obtaining an error evaluation using the error evaluation function. 若获得的误差评价指标都不在设定的阈值范围内,则需要重复上述电极消耗量的非线性预测模型建立的过程,若获得的误差评价指标在设定的阈值范围内,则,认为误差评价指标最小的非线性预测模型为最优模型,确定误差评价指标最小的非线性预测模型为电极长度检测中的电极消耗量的非线性预测模型。 If the error evaluation obtained are not in the threshold range, the need to repeat the process linear predictive model of the electrode consumption is established, if the error evaluation value obtained within the range of the set threshold, then that error evaluation index minimum linear predictive model is the best model, the linear predictive model to determine the minimum error for the linear predictive model evaluation index electrode consumption detection electrode length.

[0093] 为了更进一步的获得更为精确的电极消耗量的非线性预测模型,可以进一步选择验证样本集,对上述确定的电极长度检测中的电极消耗量的非线性预测模型进一步进行验证,比较上述非线性预测模型的输入和实际测量值的误差,如果误差在允许范围内,则确定该模型可以用于电极消耗量的在线预测;如果误差较大,分析电极消耗模型训练样本数据, 继续训练模型,可以适当增加训练模型样本数据,重复上述模型建立的过程,直至获得最优的模型。 [0093] In order to obtain the linear predictive model further more precise electrode consumption can be further selected validation sample set, linear predictive model for electrode consumption detecting the length of the electrodes is further defined to verify, compare the input linear predictive model and the actual measurement value of the error, if the error is within the allowable range, it is determined that the model can be used to predict the line electrode consumption; if the error is large, the electrode consumption model analysis of training data, the training continues model, the model may be appropriate to increase the training sample data, repeat the above process model until the best model available.

[0094] 在步骤S102,根据电极消耗量的非线性预测模型,确定当前工况下的电极消耗量。 [0094] In step S102, according to the electrode consumption linear predictive model to determine the current condition of the electrode consumption.

[0095] 在需要检测电极长度时,先获取当前工况下的电极温度T、电极功率P、炉料配比R 和电石出炉量G的参数,这些参数通常都可以实时获取,而后,利用电极消耗量的非线性预测模型,获得当前工况下的电极消耗量。 [0095] When the length of the electrodes to be detected, to obtain an electrode temperature T, the electrodes of the power P, R and the ratio of the charge amount G parameters carbide baked current operating conditions, these parameters typically are acquired in real time, and then, using the electrode consumption the amount of linear predictive model, to obtain an electrode consumption current operating conditions. 通常地,在获取这些参数之后,要进行数据的标准化处理,即进行归一化处理,使得不同量纲的数据之间可以进行运算。 Generally, after obtaining these parameters, the data to be normalized, that is normalized so that the operation can be performed between the data of different dimensions.

[0096] 在步骤S103,由当前工况下的电极消耗量,确定当前电极长度。 [0096] In step S103, the current consumption by the operating conditions of the electrode, determining a current length of the electrode.

[0097] 可以通过如下计算,通过当前工况下的电极消耗量,来确定当前电极长度H = !10+六!^+六(1*(:-六扎,1',其中:!1为当前电极长度,氏为前一时刻电极长度,六!^为电极升降量,Ad为电极单次压放量,C为压放次数,AH xhS电极消耗量,T为电极长度H和H。对应时刻的时间差。 [0097] can be calculated, the current consumption by the operating conditions of the electrode, the electrode current is determined length H = 10+ six + six ^ (* 1 (: -!! Six bar, 1 ', wherein: 1 is a! The current length of the electrode, the electrode's length for the previous time, six! ^ lift amount of the electrode, the electrode is a single pressing the Ad volume, C is the number of pressure release, AH xhS electrode consumption, T is H. H and electrode length corresponding to the time the time difference.

[0098] 这样,就由当前工况参数预测出了电极消耗量,进而由电极消耗量确定出当前电极长度,实现电极长度的软测量。 [0098] Thus, the prediction parameter from the current operating conditions the electrode consumption is determined by the electrode and thus the current consumption of the electrode length, the length of the soft-measuring electrode. 从而,可以实时监控电极的位置变化,进而指导电极的压放操作,以精确的控制各电极的压放量,使得各电极做功平衡,为电石炉的平稳生产和节能运行提供保障。 Thereby, real-time monitoring of the change in position of the electrodes, and thus the pressure release operation guide electrode, to precisely control the pressure of the volume of each electrode so that each electrode work balance, to provide protection and energy-saving operation for the smooth production of calcium carbide furnace.

[0099] 此外,本发明还提供与上述方法对应的电石炉电极长度的检测系统,参考图3所示,包括: [0099] Further, the present invention provides a furnace electrode length carbide detection system corresponding to the method described above, with reference to FIG. 3, comprising:

[0100] 电极消耗量预测模型获取模块300,用于基于模型训练样本集,获取电极消耗量的非线性预测模型,其中,电极温度T、电极功率P、炉料配比R和电石出炉量G为样本输入变量,电极消耗量为样本输出变量; [0100] electrode consumption prediction model acquiring module 300, a model based on the training sample set, acquired linear predictive model consumption electrode, wherein the electrode temperature T, the electrodes of the power P, R and the ratio of the charge amount G is baked carbide sample input variables, output variables electrode consumption of the sample;

[0101] 电极消耗量确定模块310,用于根据电极消耗量的非线性预测模型,确定当前工况下的电极消耗量; [0101] electrode consumption determination module 310, according to the linear predictive model for electrode consumption, determining the current operating conditions of the electrode consumption;

[0102] 当前电极长度确定模块320,用于由当前工况下的电极消耗量,确定当前电极长度。 [0102] The current electrode length determination module 320, the current consumed by the electrode conditions for determining the current length of the electrode.

[0103] 其中,电极消耗量预测模型获取模块300包括: [0103] wherein the electrode consumption prediction model acquiring module 300 comprises:

[0104] 通过非线性映射将模型训练样本集的输入样本映射到特征空间,构造该特征空间的最优线性回归函数; [0104] by non-linear model mapping the input training set samples are mapped into feature space, an optimal configuration of the feature space linear regression functions;

[0105] 利用结构风险最小化原则,建立最优线性回归函数的目标优化函数,在目标优化函数中,选择训练误差的2范数作为损失函数,且采用等式约束条件; [0105] With the structural risk minimization principle, the goal of establishing the optimal linear regression function optimization function, optimizing the objective function, select the training error norm as a function of loss, and the use of equality constraints;

[0106] 根据拉格朗日乘子,建立目标优化函数的拉格朗日函数; [0106] According to Lagrange multipliers, the establishment-objective optimization function Lagrangian function;

[0107] 由最小二乘法,通过拉格朗日函数确定基于拉格朗日乘子和核函数的电极消耗量的非线性预测模型。 [0107], nonlinear predictive model based on Lagrange multiplier and the electrode consumption is determined by the kernel function by the least square method Lagrange function.

[0108] 所述核函数优选为径向基核函数。 [0108] The kernel function is preferably a kernel function.

[0109] 进一步地,参考图4所示,所述模型训练样本集为多个,还包括: [0109] Further, the model training set shown in FIG. 4 is a plurality of reference, further comprising:

[0110] 误差评价模块340,基于误差训练样本集,获得电极消耗量的非线性预测模型的误差评价指标;并确定误差评价指标最小的非线性预测模型为电极长度检测中的电极消耗量的非线性预测模型。 [0110] error estimation block 340, the training sample set based on the error, the error evaluation obtained linear predictive model electrode consumption; and determining the minimum error evaluation linear predictive model is a non-electrode consumption detection electrode length linear prediction model.

[0111] 进一步地,还包括数据预处理模块330,用于将模型训练样本集进行数据预处理, 其中,电极温度T、电极功率P为当前时刻前第一预定时间范围内的平均值,炉料配比R为当前时刻前第二预定时间范围内的平均值,电石出炉量G为每批电石出炉锅数与单锅电石重量的乘积。 [0111] Furthermore, further comprising a data pre-processing module 330, the model for the training sample set data preprocessing, wherein the electrode temperature T, the electrodes of the power P as mean time before the current within a first predetermined time range, the charge R is a ratio of a second predetermined time before the current average value, G is the product of the amount of released calcium carbide, calcium carbide furnace pot batch number and a single pot by weight of calcium carbide in the time range.

[0112] 进一步地,当前电极长度确定模块320中,当前电极长度H = !10+六!^+六(1*(:-六扎,1',其中:!1为当前电极长度,氏为前一时刻电极长度,六!^为电极升降量,Ad为电极单次压放量,C为压放次数,AHxhS电极消耗量,T为电极长度H和H。对应时刻的时间差。 [0112] Furthermore, this electrode length determination module 320, the current length of the electrode ^ + H = 10+ six six (* 1 (: -!! Six bar, 1 ', wherein: the current length of the electrode 1, s is! before a time length of the electrode, six! ^ lift amount of the electrode, the electrode is a single pressing the Ad volume, C is the number of pressure release, AHxhS electrode consumption, T is H. H and electrode length corresponding to the time difference between the time.

[0113] 本说明书中的各个实施例均采用递进的方式描述,各个实施例之间相同相似的部分互相参见即可,每个实施例重点说明的都是与其他实施例的不同之处。 [0113] In the present specification, various embodiments are described in a progressive manner, similar portions of the same between the various embodiments refer to each other, are different from the embodiment and the other embodiments described each embodiment focus. 尤其,对于系统实施例而言,由于其基本相似于方法实施例,所以描述得比较简单,相关之处参见方法实施例的部分说明即可。 In particular, for embodiments of the system, since they are substantially similar to the method embodiments, the description is relatively simple, some embodiments of the methods see relevant point can be described. 以上所描述的系统实施例仅仅是示意性的,其中所述作为分离部件说明的模块或单元可以是或者也可以不是物理上分开的,作为模块或单元显示的部件可以是或者也可以不是物理单元,即可以位于一个地方,或者也可以分布到多个网络单元上。 Embodiments of the systems described above are merely illustrative, wherein the modules or units described as separated parts may be or may not be physically separated, the display member as modules or units may or may not be physical units , i.e., it may be located in one place, or may be distributed to multiple network units. 可以根据实际的需要选择其中的部分或者全部模块来实现本实施例方案的目的。 You can select some or all of the modules according to actual needs to achieve the object of the solutions of the embodiments. 本领域普通技术人员在不付出创造性劳动的情况下,即可以理解并实施。 Those of ordinary skill in the art without creative efforts, can be understood and implemented.

[0114] 以上所述仅是本发明的优选实施方式,虽然本发明已以较佳实施例披露如上,然而并非用以限定本发明。 [0114] The above are only preferred embodiments of the present invention, although the present invention has been disclosed as the preferred embodiment, however, not intended to limit the present invention. 任何熟悉本领域的技术人员,在不脱离本发明技术方案范围情况下,都可利用上述揭示的方法和技术内容对本发明技术方案作出许多可能的变动和修饰, 或修改为等同变化的等效实施例。 Any skilled in the art, without departing from the scope of the technical solution of the present invention, can take advantage of the above-described methods and technical content disclosed that many possible variations and modifications of the technical solution of the present invention, as equivalent variations or modifications equivalent embodiments example. 因此,凡是未脱离本发明技术方案的内容,依据本发明的技术实质对以上实施例所做的任何简单修改、等同变化及修饰,均仍属于本发明技术方案保护的范围内。 Thus, all without departing from the technical solutions of the present invention, any simple modification based on the technical essence of the present invention made of the above Example, equivalents, modifications and variations, provided they fall within the scope of protection of the present invention.

Claims (12)

  1. 1. 一种电石炉电极长度的检测方法,其特征在于,所述方法包括: 基于模型训练样本集,获取电极消耗量的非线性预测模型,其中,电极温度T、电极功率P、炉料配比R和电石出炉量G为样本输入变量,电极消耗量为样本输出变量; 根据电极消耗量的非线性预测模型,确定当前工况下的电极消耗量; 由当前工况下的电极消耗量,确定当前电极长度。 1. A method of detecting the length of the calcium carbide furnace electrode, wherein, said method comprising: model-based training sample set, acquired linear predictive model consumption electrode, wherein the electrode temperature T, the electrodes of the power P, charging mixing and R G is the amount of carbide baked sample input variables, output variables sample electrode consumption; linear predictive model according to the electrode consumption, the current consumption of the electrodes is determined conditions; an electrode consumption from the current operating conditions, determining current electrode length.
  2. 2. 根据权利要求1所述的检测方法,其特征在于,基于模型训练样本集,获取电极消耗量的非线性预测模型的步骤包括: 通过非线性映射将模型训练样本集的输入样本映射到特征空间,构造该特征空间的最优线性回归函数; 利用结构风险最小化原则,建立最优线性回归函数的目标优化函数,在目标优化函数中,选择训练误差的2范数作为损失函数,且采用等式约束条件; 根据拉格朗日乘子,建立目标优化函数的拉格朗日函数; 由最小二乘法,通过拉格朗日函数确定基于拉格朗日乘子和核函数的电极消耗量的非线性预测模型。 2. A detection method according to claim 1, characterized in that, based on the model training sample set, step linear predictive model obtaining an electrode consumption comprising: nonlinear mapping model input training set sample to the mapped feature space, the optimal configuration of the feature space linear regression functions; use structural risk minimization principle, the goal of establishing the optimal linear regression function optimization function, optimizing the objective function, select the training error norm as a function of loss, and the use of equality constraints; Lagrange multiplier, to establish objective Lagrangian optimization function; by the least squares method, the electrode consumption is determined based on the Lagrange multiplier and the kernel function by Lagrangian the non-linear predictive model.
  3. 3. 根据权利要求2所述的检测方法,其特征在于,所述核函数为径向基核函数。 3. The detecting method according to claim 2, wherein the kernel function is a radial basis kernel function.
  4. 4. 根据权利要求2所述的检测方法,其特征在于,所述模型训练样本集为多个,在获得电极消耗量的非线性预测模型之后,还包括: 基于误差训练样本集,获得电极消耗量的非线性预测模型的误差评价指标; 确定误差评价指标最小的非线性预测模型为电极长度检测中的电极消耗量的非线性预测模型。 4. The detection method according to claim 2, wherein said plurality of model training set, after obtaining the linear predictive model electrode consumption, further comprising: a training sample set based on an error obtained electrode consumption evaluation linear predictive model error amount; linear predictive model to determine the minimum error for the linear predictive model evaluation index electrode consumption detection electrode length.
  5. 5. 根据权利要求1-4中任一项所述的检测方法,其特征在于,所述模型训练样本集经过数据预处理,其中,电极温度T、电极功率P为当前时刻前第一预定时间范围内的平均值, 炉料配比R为当前时刻前第二预定时间范围内的平均值,电石出炉量G为每批电石出炉锅数与单锅电石重量的乘积。 The detection process as claimed in any one of the preceding claims, characterized in that the model training set preprocessed data, wherein the electrode temperature T, the electrodes of the power P of the first predetermined time before the current time average charge ratio R is in the range of a second predetermined time before the current average, the amount of released calcium carbide in the time range and the number of G baked product pot carbide single pot of each batch by weight of calcium carbide.
  6. 6. 根据权利要求1所述的检测方法,其特征在于,当前电极长度Η= !10+六!^+六(1*(:-六扎,1',其中:!1为当前电极长度,氏为前一时刻电极长度,六!^为电极升降量,Ad为电极单次压放量,C为压放次数,ΔΗΧΑ电极消耗量,Τ为电极长度Η和Η。对应时刻的时间差。 The detection method according to claim 1, characterized in that the length of the electrode current ^ Η = 10+ six + six (* 1 (: -!! Six bar, 1 ', wherein: the current length of the electrode 1,! s is the length of the electrode before a time point, six! ^ lift amount of the electrode, the electrode is a single pressing the Ad volume, C is the number of pressure release, ΔΗΧΑ electrode consumption, Τ [eta] and [eta] is the length of the electrode. time corresponding to the time difference.
  7. 7. -种电石炉电极长度的检测系统,其特征在于,包括: 电极消耗量预测模型获取模块,用于基于模型训练样本集,获取电极消耗量的非线性预测模型,其中,电极温度Τ、电极功率Ρ、炉料配比R和电石出炉量G为样本输入变量,电极消耗量为样本输出变量; 电极消耗量确定模块,用于根据电极消耗量的非线性预测模型,确定当前工况下的电极消耗量; 当前电极长度确定模块,用于由当前工况下的电极消耗量,确定当前电极长度。 7. - kind of calcium carbide furnace length of the electrode detection system, characterized by comprising: an electrode consumption prediction model acquiring means for training sample set based on the model, the electrode consumption acquired linear predictive model, wherein the electrode temperature Τ, electrode power [rho], and calcium carbide furnace charging mixing amount G R of the sample input variables, output variables sample electrode consumption; determining module electrode consumption, according to linear predictive model for electrode consumption, determining the current operating conditions electrode consumption; current electrode length determining means for consumption by the electrode current operating conditions, determining the current length of the electrode.
  8. 8. 根据权利要求7所述的检测系统,其特征在于,电极消耗量预测模型获取模块包括: 通过非线性映射将模型训练样本集的输入样本映射到特征空间,构造该特征空间的最优线性回归函数; 利用结构风险最小化原则,建立最优线性回归函数的目标优化函数,在目标优化函数中,选择训练误差的2范数作为损失函数,且采用等式约束条件; 根据拉格朗日乘子,建立目标优化函数的拉格朗日函数; 由最小二乘法,通过拉格朗日函数确定基于拉格朗日乘子和核函数的电极消耗量的非线性预测模型。 8. The detection system according to claim 7, characterized in that the electrode consumption prediction model acquiring module comprising: nonlinear mapping model input training set samples are mapped into feature space, an optimal configuration of the linear feature space regression function; use structural risk minimization principle, the goal of establishing the optimal linear regression function optimization function, optimizing the objective function, select the training error norm as a function of loss, and the use of equality constraints; Lagrange multipliers, to establish objective Lagrangian optimization function; by the least squares method, linear predictive model to determine the consumption of the electrodes and the Lagrange multiplier kernel based on Lagrange function.
  9. 9. 根据权利要求8所述的检测系统,其特征在于,所述核函数为径向基核函数。 9. The detection system of claim 8, wherein the kernel function is a radial basis kernel function.
  10. 10. 根据权利要求8所述的检测系统,其特征在于,所述模型训练样本集为多个,还包括: 误差评价模块,基于误差训练样本集,获得电极消耗量的非线性预测模型的误差评价指标;并确定误差评价指标最小的非线性预测模型为电极长度检测中的电极消耗量的非线性预测模型。 10. The detection system according to claim 8, wherein said plurality of model training set, further comprising: an error estimation module, an error based on the training sample set obtained linear predictive model error electrode consumption evaluation; and linear predictive model to determine the minimum error evaluation of electrode consumption in the detecting electrode length linear predictive model.
  11. 11. 根据权利要求7-10中任一项所述的检测系统,其特征在于,还包括数据预处理模块,用于将模型训练样本集进行数据预处理,其中,电极温度T、电极功率P为当前时刻前第一预定时间范围内的平均值,炉料配比R为当前时刻前第二预定时间范围内的平均值,电石出炉量G为每批电石出炉锅数与单锅电石重量的乘积。 11. The detection system according to claim 7-10 in claim 1, characterized by further comprising a data pre-processing module, for the model training set data preprocessing, wherein the electrode temperature T, the electrodes of the power P is before the current time of the first predetermined average charge ratio R in the time range of a second predetermined time before the current time range of the average amount G is baked carbide carbide furnace pot batch number and weight of the product calcium carbide single pot .
  12. 12. 根据权利要求1所述的检测系统,其特征在于,当前电极长度确定模块中,当前电极长度H=HQ+AHsj+Ad*C_AHxh*T,其中:Η为当前电极长度,H。 12. The detection system according to claim 1, wherein the electrode length determination module current, the current electrode length H = HQ + AHsj + Ad * C_AHxh * T, wherein: Η current electrode length, H. 为前一时刻电极长度,AHsj 为电极升降量,△d为电极单次压放量,C为压放次数,△HxhS电极消耗量,T为电极长度Η 和Η。 For the previous time length of the electrode, AHsj electrode lifting amount, △ d electrode single pressure volume, C is the number of pressure release, △ HxhS electrode consumption, T [eta] and [eta] is the length of the electrode. 对应时刻的时间差。 Time corresponding to the time difference.
CN 201510819613 2015-11-23 2015-11-23 Method and system for measuring length of calcium carbide furnace electrode CN105301985A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN 201510819613 CN105301985A (en) 2015-11-23 2015-11-23 Method and system for measuring length of calcium carbide furnace electrode

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN 201510819613 CN105301985A (en) 2015-11-23 2015-11-23 Method and system for measuring length of calcium carbide furnace electrode

Publications (1)

Publication Number Publication Date
CN105301985A true true CN105301985A (en) 2016-02-03

Family

ID=55199383

Family Applications (1)

Application Number Title Priority Date Filing Date
CN 201510819613 CN105301985A (en) 2015-11-23 2015-11-23 Method and system for measuring length of calcium carbide furnace electrode

Country Status (1)

Country Link
CN (1) CN105301985A (en)

Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US6493596B1 (en) * 1996-05-06 2002-12-10 Pavilion Technologies, Inc. Method and apparatus for controlling a non-linear mill
CN102297669A (en) * 2011-01-25 2011-12-28 东北大学 Electroslag melting furnace one kind of soft measurement of the remaining length of the electrode
CN102636624A (en) * 2012-04-28 2012-08-15 重庆大学 Method for soft measurement of alumina concentration in electrolyzer during aluminum electrolysis process

Patent Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US6493596B1 (en) * 1996-05-06 2002-12-10 Pavilion Technologies, Inc. Method and apparatus for controlling a non-linear mill
CN102297669A (en) * 2011-01-25 2011-12-28 东北大学 Electroslag melting furnace one kind of soft measurement of the remaining length of the electrode
CN102636624A (en) * 2012-04-28 2012-08-15 重庆大学 Method for soft measurement of alumina concentration in electrolyzer during aluminum electrolysis process

Non-Patent Citations (4)

* Cited by examiner, † Cited by third party
Title
何忠 等: "软测量技术在密闭电石炉生产中的应用", 《商业故事》 *
王玮: "集成多SVM的不常用备件需求预测支持系统研究", 《中国优秀硕士学位论文全文数据库信息科技辑》 *
陈刚: "软测量技术的研究及其在电石生产中的应用", 《中国优秀硕士学位论文全文数据库工程科技I辑》 *
陈寿辉: "铁合金冶炼矿热炉电极位置检测建模及优化设定研究", 《中国优秀硕士学位论文全文数据库工程科技I辑》 *

Similar Documents

Publication Publication Date Title
CN102324034A (en) Sensor-fault diagnosing method based on online prediction of least-squares support-vector machine
CN101819259A (en) Method for evaluating consistency of battery pack
Li et al. Adaptive NN output-feedback stabilization for a class of stochastic nonlinear strict-feedback systems
CN102081020A (en) Material fatigue-life predicting method based on support vector machine
CN103513640A (en) Integral optimization method and system for automatic power generation system of coal-fired unit
CN104678143A (en) Air conditioner as well as method and device for detecting power consumption of air conditioner
CN102682218A (en) Method for evaluating electricity energy efficiency of industrial user
CN102721479A (en) Online monitoring method for temperature rise of outdoor electrical device
CN103926538A (en) Variable tap-length RC equivalent circuit model and realization method based on AIC
CN102554171A (en) Breakout prediction method for continuous casting
CN101286579A (en) Method of making sets of valve controlling type accumulator by computer
CN102566551A (en) Data mining-based method for analyzing thermal power plant operation index optimal target value
CN102651116A (en) Power load data refining method
CN101968318A (en) Method for automatically controlling power of electric furnace
CN102997216A (en) Method for determining best setting value of chain furnace control system
CN105550472A (en) Prediction method of transformer winding hot-spot temperature based on neural network
CN103033762A (en) Difference value differential based lithium iron phosphate power battery power loading capacity dynamic prediction method
CN103457297A (en) Closed-loop control method for achieving active power adjustment by introducing frequency deviation feedback to have control over voltage
CN102090705A (en) Method for improving tobacco-drying process capability
CN102366967A (en) Steam curing automatic temperature control system
CN1651613A (en) Control method of carbon anode roasting production system
CN103336438A (en) Asymmetrical constrained adaptive control method of complex thermal object
CN1766130A (en) Harmonization control method for blast furnace hot blast stove system
CN101423348A (en) Integrated recognition method for sintering conditions of cement rotary kiln
CN103728879A (en) Power station boiler emission soft measuring method based on least squares support vector machine and on-line updating

Legal Events

Date Code Title Description
C06 Publication
C10 Entry into substantive examination