Month: December 2025
Difference in initial predicition for LSTM Training matlab
I have done timeseries training with LSTM in matlab. Once i am doing prediction for data apart from training i am observing difference in initial timestep then its matching quite well with actual test data. Is there any way to improve it?
Dataset has two input and one output. i have normalised dataset and havent considered time as feature in the dataset.I have done timeseries training with LSTM in matlab. Once i am doing prediction for data apart from training i am observing difference in initial timestep then its matching quite well with actual test data. Is there any way to improve it?
Dataset has two input and one output. i have normalised dataset and havent considered time as feature in the dataset. I have done timeseries training with LSTM in matlab. Once i am doing prediction for data apart from training i am observing difference in initial timestep then its matching quite well with actual test data. Is there any way to improve it?
Dataset has two input and one output. i have normalised dataset and havent considered time as feature in the dataset. deep learning, lstm MATLAB Answers — New Questions
The 3-Phase Inverter with PWM generation built using Matlab-Simulink, the output current is not same in all 3 Phases
Hello Community,
I generally work on 3D simulation tools associated with domains such as Powertrain NVH studies and I am new to Matlab-Simulink tools.
Now I am exploring the electrical/electronics domain (beginner) to get some initial data that will be required for further NVH studies.
I have made a 3-Phase Inverter with Pulse Width Modulation (PWM) circuit with help of tutorials online.
Currently I am facing this issue, where the output current extracted from the 3 Phase Inverter is not having the same magnitude(values) in
3 phases, How to correct the Matlab-simulink model regarding this issue?
Inputs given to the model,
DC Voltage source – 400 V
and Reference Sine wave blocks – Frequency 50Hz and Phase difference of 120 for 3 phases (Amplitude 1)
and carrier wave (Sawtooth generator) – Frequency 10,000 Hz (Amplitude 1)
The Load – 3-Phase Parallel RLC Branch – Only Resistor type is used with 1 Ohm resistance
Note: Regarding the grounding, two approaches I have tried – Grounding the load (RLC) and grounding the Inverter 6-switches loop (both have same results)
Thank you in advance for all the support and helpHello Community,
I generally work on 3D simulation tools associated with domains such as Powertrain NVH studies and I am new to Matlab-Simulink tools.
Now I am exploring the electrical/electronics domain (beginner) to get some initial data that will be required for further NVH studies.
I have made a 3-Phase Inverter with Pulse Width Modulation (PWM) circuit with help of tutorials online.
Currently I am facing this issue, where the output current extracted from the 3 Phase Inverter is not having the same magnitude(values) in
3 phases, How to correct the Matlab-simulink model regarding this issue?
Inputs given to the model,
DC Voltage source – 400 V
and Reference Sine wave blocks – Frequency 50Hz and Phase difference of 120 for 3 phases (Amplitude 1)
and carrier wave (Sawtooth generator) – Frequency 10,000 Hz (Amplitude 1)
The Load – 3-Phase Parallel RLC Branch – Only Resistor type is used with 1 Ohm resistance
Note: Regarding the grounding, two approaches I have tried – Grounding the load (RLC) and grounding the Inverter 6-switches loop (both have same results)
Thank you in advance for all the support and help Hello Community,
I generally work on 3D simulation tools associated with domains such as Powertrain NVH studies and I am new to Matlab-Simulink tools.
Now I am exploring the electrical/electronics domain (beginner) to get some initial data that will be required for further NVH studies.
I have made a 3-Phase Inverter with Pulse Width Modulation (PWM) circuit with help of tutorials online.
Currently I am facing this issue, where the output current extracted from the 3 Phase Inverter is not having the same magnitude(values) in
3 phases, How to correct the Matlab-simulink model regarding this issue?
Inputs given to the model,
DC Voltage source – 400 V
and Reference Sine wave blocks – Frequency 50Hz and Phase difference of 120 for 3 phases (Amplitude 1)
and carrier wave (Sawtooth generator) – Frequency 10,000 Hz (Amplitude 1)
The Load – 3-Phase Parallel RLC Branch – Only Resistor type is used with 1 Ohm resistance
Note: Regarding the grounding, two approaches I have tried – Grounding the load (RLC) and grounding the Inverter 6-switches loop (both have same results)
Thank you in advance for all the support and help three-phase inverters, pulse width modulation (pwm), simulink MATLAB Answers — New Questions
Closing the Property Inspector in R2025b
The Property Inspector pane does not appear to have a ‘Close’ button in R2025b, nor does its sidebar icon appear to have a dropdown selection to close, unlike most other sidebar icons. Why is that? How is it intended for the Property Inspector to be closed?The Property Inspector pane does not appear to have a ‘Close’ button in R2025b, nor does its sidebar icon appear to have a dropdown selection to close, unlike most other sidebar icons. Why is that? How is it intended for the Property Inspector to be closed? The Property Inspector pane does not appear to have a ‘Close’ button in R2025b, nor does its sidebar icon appear to have a dropdown selection to close, unlike most other sidebar icons. Why is that? How is it intended for the Property Inspector to be closed? graphics, property inpsector MATLAB Answers — New Questions
Office 365 for IT Pros December 2025 Update
Monthly Update #126 Now Available for Download
The Office 365 for IT Pros team is happy to announce the availability of the December 2025 update for the Office 365 for IT Pros (2026 edition) eBook. This is the 126th monthly update. We’ve also published an update for the Automating Microsoft 365 with PowerShell eBook (now at version 18.1).
Subscribers can download the updated PDF and EPUB files from Gumroad.com. The link in the receipt you received always fetches the latest files. See our FAQ for more information about downloading updated, and our change log for details of what’s changed in update #125.
The Agent Age Unfolds
In November, I had the opportunity to attend the Microsoft Ignite conference in San Francisco. Much as I like San Francisco, I was unimpressed by Ignite 2025. Much to my amazement, Microsoft announced that the 2026 Ignite event will also be in San Francisco, when I suspect that the conference will be dominated again by AI, Copilot, and Agents.
One of the big Microsoft announcements at Ignite 2025 was Agent 365, the new “control plane” for agents. It would be a grave mistake to imagine that Agent 365 is the universal panacea for agent management. Like most Microsoft management initiatives, Agent 365 is likely to be full of holes and incomplete. At least, ISVs hope that this truth will endure because that’s how they make their money. Just look at the market that’s sprung up in the last two years around Entra ID app management.
Agent 365 has legitimized the topic of agent management within Microsoft 365 tenants. At this point, it seems likely that most if not all tenants will have some element of agent activity within a few short years. However, it’s not clear what the scope of the agents will be, who will build them, what management is required to ensure that agents are secure, compliant, and protected against attack.
Possibly even more important, no one knows how much AI and agent processing will cost. The Microsoft 365 Copilot license is expensive, but at this point it’s the simplest and easiest way to manage AI licensing because of what the Copilot license includes. If Microsoft doesn’t simplify and clarify how they will charge customers for agent processing, pay as you go agent licensing, consumption, measurement, and reporting is likely going to be a core skill for tenant administrators to acquire in the future.
It will be interesting to see how the Agent 365 control system evolves over the coming years to deal with these and other challenges. It will be even more interesting to observe what gaps ISVs find in Agent 365 to exploit and complement with what I expect to be a host of agent management solutions.
Agent Management and the Office 365 for IT Pros eBook
In the meantime, the Office 365 for IT Pros team is thinking about how to incorporate agent management into the book. Our book is all about excellence in Microsoft 365 tenant administration with special focus on core workloads like Entra ID, Exchange Online, SharePoint Online, and Teams. I suspect that the line we will take will focus on core agent management (like the Agent 365 section of the Microsoft 365 admin center) and how to automate administrative tasks with agents. Expanding past these boundaries would require a separate book. Maybe we’ll do that too, but that’s another day’s work.
One thing’s for sure. We have many announcements but much of what’s been announced remains to be exposed in public. When that happens, we’ll start to build content. Expect to see the results of that work in the 2027 edition.
On to Update #127
But we have more pressing matters on our plate, like continuing to roll out the monthly updates for the 2026 edition. Even though not much happens in Microsoft in December, we’ll product update #127 in early January 2026. The fun continues!
Pinning the AxesToolbar (revisited)
The AxesToolbar was introduced about 6 years ago,
However, I never embraced it, because it was not possible (at least not at the time) to pin the toolbar so that the exploration buttons are perpetually in view. A pinned view of the buttons is something I prefer from the days when the buttons were on the figure toolbar, though I do see the logic of tying the buttons to an axes, rather than a figure.
I thought I would check back in and see if this pinning capability now exists. I seem to remember I was not the only one calling for this feature. So is there now a way to make the AxesToolbar perpetually visible, even when the mouse is not hovering over the axes?The AxesToolbar was introduced about 6 years ago,
However, I never embraced it, because it was not possible (at least not at the time) to pin the toolbar so that the exploration buttons are perpetually in view. A pinned view of the buttons is something I prefer from the days when the buttons were on the figure toolbar, though I do see the logic of tying the buttons to an axes, rather than a figure.
I thought I would check back in and see if this pinning capability now exists. I seem to remember I was not the only one calling for this feature. So is there now a way to make the AxesToolbar perpetually visible, even when the mouse is not hovering over the axes? The AxesToolbar was introduced about 6 years ago,
However, I never embraced it, because it was not possible (at least not at the time) to pin the toolbar so that the exploration buttons are perpetually in view. A pinned view of the buttons is something I prefer from the days when the buttons were on the figure toolbar, though I do see the logic of tying the buttons to an axes, rather than a figure.
I thought I would check back in and see if this pinning capability now exists. I seem to remember I was not the only one calling for this feature. So is there now a way to make the AxesToolbar perpetually visible, even when the mouse is not hovering over the axes? axestoolbar, navigation, exploration, graphics, axes, plots, axtoolbar MATLAB Answers — New Questions
using knob to adjust audio frequency
How can I use a knob to adjust the frequency of audio such as the handel.mat soundHow can I use a knob to adjust the frequency of audio such as the handel.mat sound How can I use a knob to adjust the frequency of audio such as the handel.mat sound appdesigner, knob, audio MATLAB Answers — New Questions
I want to combine these two plots so they are placed sequentially in a bigger graph spanning -2 to 2. everything i try alters the data the plots show which i want to avoid.
Post Content Post Content fea, graph MATLAB Answers — New Questions
LQR Controller for Nonlinear System Tracking with Time-Varying Linearization
Hello everyone,
I’m trying to implement an LQR controller for a nonlinear system to track a reference trajectory. I need to linearize around each point of the trajectory and compute time-varying LQR gains.
My current problem is how to implement the following:
From a nonlinear system: xdot = f(x, u)
I need to linearize around each reference point (x_ref(t), u_ref(t)) to get A(t) and B(t)
Compute LQR gains K(t) at each time step
Apply the control input
My MATLAB code is as follows:
matlab
Q = zeros(size(A));
R = eye(size(B,2));
[K, S, e] = lqr(A, B, Q, R);
A_cl = A – B*K;
B_cl = zeros(size(B));
C_cl = C;
D_cl = D;
sys_cl = ss(A_cl, B_cl, C_cl, D_cl);
[y_cl_dev, t_sim, x_cl_dev] = initial(sys_cl, delta_x0, t);
x_cl_phys = x_cl_dev + repmat(X_op.’, length(t_sim), 1);
y_cl_phys = y_cl_dev + repmat((C*X_op).’, length(t_sim), 1);
My Question:
How can I use ODE45 to, at each point, linearize the system to get A and B, compute K, apply the control input, and then obtain xdot and y?
How convert the
initial
function to a loop-based simulation.
thanks.Hello everyone,
I’m trying to implement an LQR controller for a nonlinear system to track a reference trajectory. I need to linearize around each point of the trajectory and compute time-varying LQR gains.
My current problem is how to implement the following:
From a nonlinear system: xdot = f(x, u)
I need to linearize around each reference point (x_ref(t), u_ref(t)) to get A(t) and B(t)
Compute LQR gains K(t) at each time step
Apply the control input
My MATLAB code is as follows:
matlab
Q = zeros(size(A));
R = eye(size(B,2));
[K, S, e] = lqr(A, B, Q, R);
A_cl = A – B*K;
B_cl = zeros(size(B));
C_cl = C;
D_cl = D;
sys_cl = ss(A_cl, B_cl, C_cl, D_cl);
[y_cl_dev, t_sim, x_cl_dev] = initial(sys_cl, delta_x0, t);
x_cl_phys = x_cl_dev + repmat(X_op.’, length(t_sim), 1);
y_cl_phys = y_cl_dev + repmat((C*X_op).’, length(t_sim), 1);
My Question:
How can I use ODE45 to, at each point, linearize the system to get A and B, compute K, apply the control input, and then obtain xdot and y?
How convert the
initial
function to a loop-based simulation.
thanks. Hello everyone,
I’m trying to implement an LQR controller for a nonlinear system to track a reference trajectory. I need to linearize around each point of the trajectory and compute time-varying LQR gains.
My current problem is how to implement the following:
From a nonlinear system: xdot = f(x, u)
I need to linearize around each reference point (x_ref(t), u_ref(t)) to get A(t) and B(t)
Compute LQR gains K(t) at each time step
Apply the control input
My MATLAB code is as follows:
matlab
Q = zeros(size(A));
R = eye(size(B,2));
[K, S, e] = lqr(A, B, Q, R);
A_cl = A – B*K;
B_cl = zeros(size(B));
C_cl = C;
D_cl = D;
sys_cl = ss(A_cl, B_cl, C_cl, D_cl);
[y_cl_dev, t_sim, x_cl_dev] = initial(sys_cl, delta_x0, t);
x_cl_phys = x_cl_dev + repmat(X_op.’, length(t_sim), 1);
y_cl_phys = y_cl_dev + repmat((C*X_op).’, length(t_sim), 1);
My Question:
How can I use ODE45 to, at each point, linearize the system to get A and B, compute K, apply the control input, and then obtain xdot and y?
How convert the
initial
function to a loop-based simulation.
thanks. lqr, mpc, control, pid MATLAB Answers — New Questions
