好的，练习通过校准液位传感器读取储罐的反馈。
OK the exercise read the feedback off tank by calibrating level sensor.

我们已经在上一个练习中完成了这一点，因此该代码已经存在，使用它生成比例控制算法。
This we have already done in the last exercise so this code is already there generate proportional control algorithm using a it.

那么让我们看看这就是我们的西门子TIAA。
So let's see that this is our Siemens TIAA.

现在，如果您看到这里的代码，我想为您放大。
Now if you see the code here I want to zoom in for you.

好的。
All right.

显示这个。
Display this.

你知道我们以前已经做到了。
You know that we have done it before.

这是买的绝对值减去水平，不买的减去水平就是误差。
This is the absolute value bought minus level not bought minus level is the error.

你的输入水平是多少，就是反馈。
What is your input level is the feedback.

因此，如果您回到演示部分，这就是设定点，您的输入减去反馈就是级别。
So if you go back to a presentation part this is the set point is your input minus feedback is the level.

所以这是输入减去反馈设定值显示。
So this is input minus feedback setpoint display.

我们知道，如果您去工厂，则已经设定点显示。
This we know that already set point display is if you go to factory.

如果您看到这里，这就是设定值显示。
And if you see here this is the setpoints display.

这是由 pod 直接显示的。
This is being displayed by the pod directly.

因此，无论您拥有什么值，无论您在设定点中拥有什么值，这都会真正进入设定点显示。
So whatever value you have is whatever value you have in the set point this would really go to the set point display.

所以我们只是在这个数字显示器上显示这个点值。
So we are just displaying this point value in this numeric numeric display.

现在，级别再次直接通过级别显示 级别显示位于此处，并且来自级别计。
Now level displays again directly by level level display is here and this comes at a level meter.

所以这被赋予了二元值。
So this has been given dyadic values.

那你有什么。
Then what do you have.

这是上次的新内容。
This is something new from the last go.

暂时地。
Temporarily.

好吧，我要做的是，如果你看到我工厂的这一部分，我会讨论这个，因为我正在使用工厂，增益值如何从 0 到 10 好吧，这意味着如果我得到超过 10 的增益直接使用我的发射器，但我想将增益提高到 75%。
OK what I'm going to do is if you see this part of my factory I'll discuss this because I'm using factory how the gain has value from zero to 10 OK which means I get out to gain more than 10 if I directly use my emitter but I want to increase the gain up to 75 percent.

好吧，我不想要只有 10 个。
OK I don't want just 10.

所以我所做的我使用了一个常数七点五，这意味着如果我的环境已经死了，临时值将变为 1 5 的倍数到 10，即 75 K，我将这个实际值引用为双倍，因为我想要以展示我的胜利。
So what I did I used a constant seven point five which means if my is dead from my environment the value in temporary will go multiple of 1 5 into 10 which is 75 K and I'm quoting this real value into double and because I want to display in my victory.

那我该怎么办呢。
So what do I do.

另一个变量这个 PKP Despard 这是我的。
Another variable this PKP Despard This is my.

这是我的输出变量。
This is my output variable.

这是从我的临时变量 10 Dampierre 获取值。
This is getting the value from 10 Dampierre is my temporary variable.

这是我的建议。
Here is my tip.

好的，这只是首先计算它，然后将其移至 K.P.
OK this is just to first calculate it and then move it over to K.P.

展示。
display.

好的，我也可以在这里直接使用它，但这只是为了向您展示您也可以真正使用现在所谓的错误。
Ok I can also use this directly here but this is just to show you that you can also use really what's now called the error.

误差为点负电平。
The error is spot minus level.

这里又是同样的事情。
Again this is the same thing here.

这就是错误。
This is the error.

我也可以在这里使用这个错误，这将是同样的事情。
I can also use this very error here that would be the same thing.

现在失败的是输出。
Now failing that is the output.

所以也许我可以给你写一些评论。
So maybe I can write it for you some comment.

输出等于误差乘以增益，我可以再次写出不成比例的增益转义，你显示这个，因为我想将乘以一个常数，因为我需要更多的增益。
Output is equal to error multiply by gain and I can write again not proportional gain escape you display this one because I want to have keep the multiply by a constant because I need to have more gain.

我想展示这个增益限制的增益从零增加到 10 到七十五。
I want to show the gain of this gain limit increases to limit increases from zero to 10 to you to seventy five.

这样您就可以看到控制器的响应。
So you can see the response of a controller.

但收益从零到 75。
But the gains zero to 75.

如果我不使用这个并且是 5，我只能看到从 0 到 10 的响应，这不是一个好主意，因为当您研究推进控制器时，您应该看到增益高达 75 的效果。
If I do not use this and been five I can only see the response from zero to 10 which is not a good idea because when you do study the propulsion controller you should see the effect of gain up to 75.

所以这就是为什么我选择了7 5。
So that's why I took the 7 5.

如果您在这里取 5 个，那么您可以看到从 0 到 50 的增益。
If you take five here then you can see the gain from zero to 50.

所以这是简单的数学。
So it's simple mathematics.

所以我的输出实际上是误差乘以增益。
So my output is actually error multiply by gain.

好的。
OK.

因此，让我们看看未知响应消耗输出与消耗输入的消耗程度如何，这意味着这意味着这在板上正在缓和，消耗输出基本上是我的声音。
So let's see how well the unknown response draining output is draining equal to draining input which means which means this is moderating in board and draining output is basically my voice.

这是我心力交瘁的沃伊尔斯。
This is my draining Voyles.

所以这与我在这里给这个旋钮的值成正比。
So this is directly proportional to what what value I give to this knob here.

因此，如果您对这种联系有任何疑问，并且这个变量只是流行，那么您来我会回复您。
So if you have any doubt in this connections and this variable is just pop you come I'll get back to you.

现在让我们在订单控制器中下载此逻辑并查看响应。
So let's download this logic now in order controller and see the response.

加载配置，我们就完成了。
Load the configuration and we are done.

那么，让我们去我们的三人组去玩吧。
So let's go to our trio of go to play.

现在让我们获取响应，然后增益为 1。
Now let's take the response then the gain is one.

好的。
OK.

所以收益是一。
So the gain is one.

现在让我们看看对某个设定点的响应。
Now let's see the response will some setpoint.

你可以看到这是 2 1 6 8。
You can see that this is 2 1 6 8.

而且这个平均值是成正比的，你的同胞有相同的值填充就会等于误差。
And this is the average is proportional to your fellow has the same value of filling will be equal to error.

好的，让我们添加一些距离。
OK so let's add some distances.

这是增益为 1 的比例控制。
Now this is a proportional control in which the gain is 1.

因此球会对错误做出反应，因为错误是输出等于编辑，而增益是一旦填充输出等于错误。
So the ball will react to the error because the error is the output is equal to edit and again gain is once a filling output is equal to error.

因此，如果你把等于误差的东西去掉，你甚至会在某个时刻看到你会出现稳态误差。
So if you take out what is equal into error you even see at one moment of time you will have a steady state error.

好吧，你会注意到这里。
OK you will notice that here.

现在你的等级是 1 x 9。
Now your level is 1 by 9.

它正在增加。
It's increasing.

而且你在这里也接受了一些培训。
And you also have some training here.

吃得不足是指 23% 未吃完，填充是指 14% 是未吃。
Undereating is 23 percent open filling is 14 percent open.

因为这种不匹配不是不匹配而是不是一个好的算法。
And because of this mismatch not mismatch but not a good algorithm.

我们有这个持续的错误。
We have this constant error.

现在我们会说我们是。
Now we will say are we.

为啥不增加增益呢。
Why don't you increase the gain.

当然，我们可以。
Of course we can.

因此，让我增加一些更多的干扰，让我们来做这个吧。
So let me increase some more disturbance end badly let's let's do this one.

你知道我们的关系非常紧凑。
You know we're pretty compact.

所以这已经是它的顶峰并重新开始了。
So this is already on its top it and start again.

所以你可以看到这条蓝线和屏幕上的蓝线就是设定点。
So you can see that this blue line and the screen line this is the blue line is the set point.

这条绿线是反馈。
This green line is the feedback.

你有一个稳态误差，这对两个人来说太多了。
You have a steady state error and that's too much for two.

现在让我们看看当我增加增益时的响应。
Now let's see the response when I increase the gain.

好的，我将把增益增加到二到三，假设增加了三个增益，您会注意到误差将从这里减少，控制器试图消除误差，但因为比例控制器只检查当前误差不检查最后一个错误。
Ok so I'm going to increase the gain to two to three let's say three of the gain is increased and you will notice the error will be reduced from here controller trying to remove the error but because the proportional controller only checks the current error it does not check the last error.

你仍然会保持稳定的状态。
You will still have a steady state.

你在这里看到的已经减少的东西是减少了，你说嘿罗杰为什么不增加更多的增益。
And what this has been reduced you see here it is reduce and you say Hey Roger why don't you increase the gain more.

也许我们最好这样做。
Maybe we better do this.

好吧，让我们尝试一下。
All right let's try that.

因此，我们将游戏增加到 6 个。
So let's increase the game to let's say 6.

当然，这正在减少。
So of course this is reducing.

您还可以在这里看到这一点，因为现在输出是 6 乘以误差。
And you can also see that here because now the output is output as 6 multiply by the error.

您可以看到，Obert 在 6 个月后的 12 点 5 分近似为 3。
And you can see that Obert a six month twelve point five is proximately 3.

所以我们在三点感觉奥卢斯。
So we're feeling Aulus at three.

但除此之外还有 4.4。
But besides this again a 4.4.

现在让我们进一步增加误差。
Now let's increase the error more.

对不起。
Sorry.

KPK 可以多赢 10 乘 10。
The KPK can gain more 10 by 10.

如果您看到此错误，您会看到所有更改与该错误成比例填充。
If you see this error you see the proportional fill of all changes with the error.

现在误差已减小到增益 0.3，但仍然存在误差。
Now error has been reduced to a gain 0.3 but still there is an error.

好的。
OK.

好的。
All right.

10 点到 15 点我们还会再来的。
We're going to be getting again again that's 10 to 15.

现在你已经注意到了。
Now you've noticed.

你知道，当你继续增加误差时，尽管当你继续增加增益时，EDID 会减小，但在某些时候你会发现你会发现振荡。
You know when you keep increasing the error although the error when you keep on increasing the gain although the EDID will reduce but at some point you will find you will find oscillations.

现在我们只是通过持续的干扰来监测响应。
Right now we're just monitoring the response with the constant disturbance.

于是让让扰乱，看看反应。
So Jean the disturbance and see the response.

但无论如何我们看到误差已经减少了0.2。
But anyways we have seen that the error has been reduced 0.2.

因此，让我们将增益增加 25。
So let's increase the gain more 25.

所以25倍的错误就是你填的是1。
So 25 multiple of error is you are filling one.

因此您会注意到错误已经被使用了 0.1 桶。
So you will notice the error has been in use for either by 0.1 keg.

让我们将其增加到 30。
Let's increase it more up to 30.

现在这里的错误已经被使用得更多了。
Now here the error has been had used more.

这是百分之一的错误。
It's 1 percent error.

所以这是相当不错的。
So it is quite good.

相当好的反响。
Quite good response.

因此，现在我们可以更进一步，将错误减少到零。
So let's even go more now the error has been reduced to zero mind.

八点九。
Eight point nine.

这样我们就可以反复查找错误就OK了。
So we can go more and over find the error to grow OK.

现在我们的值为 46，您可以看到误差几乎为 3.5%。
So right now we are at 46 and you can see the error is almost 3.5 percent.

但因为现在误差非常敏感，并且由于增益较高，这种情况正在发生变化。
But because the error is very sensitive now and because this is changing because of the higher gain.

我的感觉是低点仍然有很大差异，所以这是进行大量比赛的缺点。
My feeling was the lows still vary a lot so this is the disadvantage of having a lot of game.

这也是非常感人非常感人的错误。
It's also very moving very moving the error.

但有时我们会出现振荡。
But sometimes we get oscillations.

这被称为振荡，因为场墙会不断地打开和关闭，场墙会变得疯狂。
This is called oscillations because the field wall goes on and off the wall goes crazy.

因此，如果按比例填充以在增益处进入 BB，您就会知道会破坏系统。
So if the filling is done so proportionally to enter the BB at the gain and it will you know destroy the system.

现在我听说使用增益，但到目前为止你看不到任何解决方案。
Now I have heard use the gain and so far you can see no solution has been.

所以，如果你看到这里的回应，这不是设计的，这是太多的隔离。
So if you see the see the response here this is this is not designed this is so much isolations.

好的。
OK.

因此，一方面，如果您注意到这里的比例增益，我们总是有一个稳态误差。
So in one hand if you notice here the proportional gain we always have a steady state error.

如果你增加博弈来减少稳态误差，你就会出现振荡，这就是现在的一点。
This is the one point now if if you increase the game to reduce that steady state error you will get oscillations.

所以你不能进行权衡，你可以得到稳定的响应和更少的错误。
So you can't have a trade off very you can have a stable response with less error.

所以一般来说你不会看得太高收益以避免振荡。
So generally you don't look too much high gains to avoid oscillations.

我们只是采取较少的增益，以便我们通过稳定的设定误差（这就是这个）进行妥协。
We just take less gain so that we compromise by having a steady set error which is this one.

如果这个过程让您高兴，您可以隐藏推进控制器。
If that process makes you happy you can go hide what propulsion controller.

否则补救措施是使用 indic。
Otherwise the remedy is to use indic.

控制。
control.

好的。
OK.

让我们通过更改 tomans 来研究响应，我已将可放电量增加到九点八。
Let's study the response by changing the tomans I have increased the dischargeable to nine point eight.

如果您在此处看到 0.3，则 98% 的初始状态是开放的。
98 percent to start with is open right now if you see the 0.3 in here.

这也是错误。
Also this is the error.

因此，让我们将设定点设置为 5.5 6，此时由于 K.P.，它已百分百正确。
So let's the set point set point is 5.5 6 at this point it's hundred percent on because of K.P.

和错误。
and error.

一旦输入值降低，就知道 70% 到 80% 的评级就可以了。
And as soon as the enter decreases the value of a decrease in it's known is that 70 to 80 percent rating is OK.

你知道，看到这里的回应，它试图在编辑阅读时赶上设定点，并避免因为游戏而做出有价值的评论，让我们将其设为 15％，这样你就可以看到，因为感觉的比例意义关系接近八点八点，你仍然有五点的空气偏差，即百分之五，这是不可接受的。
You know see the response here it's trying to catch up the set point at one moment the editors read using and avoid valuable reviews because of the game let's make it 15 percent so you can see because of the proportional meaning relation felt was close to eight point eight and you still have an air off point five which is five percent which is not acceptable.

好的。
OK.

这是关于比例控制的。
So this is about a proportional control.

如果您注意到这里，您会注意到这里，如果我们总结了我们必须粘贴响应的所有内容，我将在本次讲座之后粘贴响应，但在这里您会看到优势。
And if you if you notice here you notice here if we sum up everything we have to paste a response I will paste the response after this lecture but here you see the advantage.

当然，这是非常简单且最简单的方法。
It's it's quite simple and easiest way of course.

这很简单，也很容易，因为我们只需要这样做。
It's simple and it's easy because we just have to.

你只能得到帕达姆它或保持现状。
You get only Padam it or keep to be.

并且它提供了良好的稳定性。
And it provides good stability good stability.

如果我们优化了游戏优化增益就意味着增益不会太大。
If we have optimized game optimized gain means not too much gain.

如果您的收益太多，我们就可以介入。
Then we have good step into if your gain is too much.

那么你可能没有稳定性，因为它会振荡缺点说它有偏移。
Then you may not have the stability because it would oscillate disadvantage said it has offset.

我们已经看到了。
We have seen it.

稳定状态与期望值之间有什么区别？
What is the difference between a steady state between and the desired value.

所以主要的缺点就是偏移。
So the main disadvantage is just the offset.

我们必须你必须删除这个偏移量。
We have to you have to remove this offset.

这样它将切换到一个新的控制器，即 P I can k。
So that it will switch to a new controller which is a P I can k.

因此，如果您对此控制器有任何疑问，可以向我发送一条命令，而最新要做的就是缺点。
So if you have any doubt in this in this controller you can pop me a command and the latest one to do is disadvantages.

它有一个稳定的状态。
It has a steady state.

因此，我将把这个屏幕从幻灯片上粘贴到幻灯片上，并附上一张图表。
So I'm going to paste this screen off the slide down the slide with a graph.

好的，谢谢您的观看。
All right thank you for watching.

我们将在下一个视频中见到您。
I'll see you in the next video.