Spanish Guide + PDF + ready-to-import flow.json for Heishamon Dashboard (2026)

[reivax2]

Hi everyone,
Does anyone have a complete .json flow (ready to import) for HeishaMon / Panasonic Aquarea control that I can copy and paste directly into Node-RED?
I'm looking for all the gauges (COP, consumption, temperatures, defrost, etc.).
Thanks in advance!

[edterbak]

I dont mind you giving a link to your page... once.
You probably did a good job with your manuals.

But please do not spam each issue in this github page. I myself have never done that. Never had the need to do so. I find it is not so polite. I Count now 4 times you direct people to your own website where people can pay for manuals or something.
All nice and Congratulations on your work...
But, Please stick to this thread is you want to say something about your work. That is fine with me.

I will remove all other posts with links to your website in other threads. So please stay on topic.

[edterbak]

https://payhip.com/GuiasAerotermiaYsolar

[edterbak]

@reivax2
You can add screenshot and information about what it can do for the user here. But please contain it to this issue. 😄

[reivax2]

@edterbak thank you for the previous warning and for giving me the opportunity to clarify.

Here are some screenshots of the guide I mentioned (2026 version for HeishaMon V5 + local control Panasonic Aquarea):

What it provides to the end user:

• Step-by-step configuration for HeishaMon V5 on Panasonic models J/K/L/M.
• Ready-to-import flow.json in 1 click (local dashboard with real-time visualization of COP, kWh, heating curve).
• Solves frequent problems seen in forums (mis-mapped sensors, import failures, version compatibility issues).
• Includes 58 pages with diagrams, checklists and clear explanations in Spanish.
• Focused on real installations (small apartments/houses, no cloud dependency).

If you think it could be useful for the community, I can prepare a summary or basic section (without promotion) to include in the repo documentation or gallery.

Thanks for all

[edterbak]

Oke, interesting.

Can you explain a little on what your dashboard is offering compared to the dashboard I am offering? What are the key differences?
Why and when should a user use the flow from my github?
Why and when should a user use the flow from your page?

Regards,

[edterbak]

@reivax2
Can you still respond to my question? Im curious

[reivax2]

Hi edterbak,

Thank you for your continued interest. As a multi-technical industrial installer with years of real-world testing experience in my home, my Node-RED workflow focuses on maximum efficiency:

• Advanced and ultra-lightweight Node-RED for fast processing.

• Real-time calculations with AI (Grok, Perplexity, Manu) for optimal compensation curves (better winter COP/summer EER).

• Predictive local weather API (Spain) for proactive management.

• Modulation logic with calculated offset and daily cycles.

The open-source code is about 70% functional (it still needs to integrate the local API and fully activate the AI, which requires personal contact). Both approaches are valuable depending on the needs

If I've bothered you, my apologies. I'll leave the topic here and reduce my activity on this forum.

Best regards.

[edterbak]

@reivax2
I am truely sorry, but I still have no clue what you are saying and what it is you have built.

Your reply gives me a very very high AI vibe, a lot of buzzwords without real content.
"Multi-technical installer with years of real world testing experience in my home"? I think this is the fluff definition of any home developer...

"Advanced and ultra-lightweight node red for fast processing"? Node Red is Node Red... nothing fancy or ultra light weight about it. Maintained by the Node Red development team.. They did a good job, but there is nothing you can do to make it lightweight.

"Real time calculations for compensation curves".... Why would you need realtime calculations with AI? The Curve is the curve, you can move it due to the seasons, but do you want AI for this purpose ? Is it realtime? If you are using AI for this, how light-weight is this solution then? Again raising more questions in my mind...

The only thing I read which starts to make (little) sense, is the predictive weather API interaction. But that too could just be fetching weather information via API and using the 1hour 1day ahead predictions in a flow

Of course, I could be completely wrong..
But from what I see here,
An AI generated web page, AI images, AI texts
In this github response, I think you are also using AI.

I could read this as a lot of red flags for a scam operation. To get people to pay for nothing using fancy AI generated pages.

If I do not get any more clarification on this and if you have not convinced me that this is legit and safe, I will remove the posts fully. I do not want to be responsible (or helped to facilitate) for people to get ripped off.

I hope I am wrong. If so, please make me understand.

Regards,

[reivax2]

Hello, I completely respect your comment and opinion.
I know exactly what I programmed and why I created it. If you don't believe me, that's fine too.
Below, you'll find an example of the code and its explanation.
This code will flag errors because there's no APY and IP or specific values ​​for your Heishamon.
From here, you can do whatever you want, but don't use my code to expand or update your control panel.

FLUJO NODE-RED COPY AND PASTE (nodered import)

{"id":"aerotermia_ia","type":"tab","label":"Aerotermia IA + AEMET TU CIUDAD","disabled":false,"info":"Cambia CODIGO_INE por tu municipio AEMET"},
{"id":"api_key","type":"ui_text_input","z":"aerotermia_ia","name":"AEMET API Key","label":"API Key AEMET (opendata.aemet.es)","group":"dashboard","width":"6","x":180,"y":60,"wires":[["save_key"]]},
{"id":"save_key","type":"function","z":"aerotermia_ia","name":"Guardar Key","func":"flow.set('aemet_key',msg.payload); return {payload:'Key guardada - Listo para IA'};","x":420,"y":60,"wires":[["status"]]},
{"id":"trigger","type":"inject","z":"aerotermia_ia","name":"CADA 20 SEGUNDOS","props":[{"p":"payload","v":"","vt":"date"}],"repeat":"20","once":true,"x":180,"y":100,"wires":[["aemet_req"]]},
{"id":"aemet_req","type":"function","z":"aerotermia_ia","name":"AEMET TU CIUDAD","func":"let key=flow.get('aemet_key')||'TU_API_KEY';\nlet CODIGO_INE='08086'; // CAMBIA POR TU MUNICIPIO AEMET\nlet fecha=new Date().toISOString().split('T')[0];\n\nif(key==='TU_API_KEY'){return [{payload:{error:'Ingresa API Key'}}];}\n\nmsg.url=[https://opendata.aemet.es/opendata/api/prediccion/especifica/municipio_horaria/${CODIGO_INE}?fecha=${fecha}&api_key=${key};\nmsg.method='GET';msg.headers={'Accept':'application/json'};\nreturn msg;","x":450,"y":100,"wires":[["http_aemet"]]},
{"id":"http_aemet","type":"http request","z":"aerotermia_ia","name":"AEMET API","ret":"obj","x":680,"y":100,"wires":[["parse_aemet"]]},
{"id":"parse_aemet","type":"json","z":"aerotermia_ia","name":"Parse AEMET","x":860,"y":100,"wires":[["ia_calc"]]},
{"id":"heishamon","type":"mqtt in","z":"aerotermia_ia","name":"HeishaMon TOPs","topic":"heishamon/top/#","qos":"2","broker":"heishamon_broker","x":180,"y":160,"wires":[["parse_top"]]},
{"id":"parse_top","type":"function","z":"aerotermia_ia","name":"Parse TOP6/27/114","func":"let t=msg.topic.split('/').pop();let v=parseFloat(msg.payload);\nif(isNaN(v)) return null;\nswitch(t){\n case '6': flow.set('top6',v); break; // T_impulsion\n case '27': flow.set('top27',v); break; // T_exterior\n case '114': flow.set('cop',v); break; // COP\n}\nreturn null;","x":420,"y":160,"wires":[]},
{"id":"ia_calc","type":"function","z":"aerotermia_ia","name":"IA CURVA OPTIMA (20s)","func":"let t_ext=flow.get('top27')||msg.payload?.temp||10;\nlet t_imp=flow.get('top6')||45; let cop=flow.get('cop')||4;\n\n// Historicos IA (aprende tu sistema)\nlet hist_ext=context.get('hist_ext')||[]; let hist_imp=context.get('hist_imp')||[];\nhist_ext.push(t_ext); hist_imp.push(t_imp);\nif(hist_ext.length>100){hist_ext.shift();hist_imp.shift();}\ncontext.set('hist_ext',hist_ext); context.set('hist_imp',hist_imp);\n\nlet n=hist_ext.length;\nif(n>20){\n // REGRESION LINEAL: Tu curva personalizada\n let sum_x=hist_ext.reduce((a,b)=>a+b,0), sum_y=hist_imp.reduce((a,b)=>a+b,0);\n let sum_xy=hist_ext.reduce((a,b,i)=>a+bhist_imp[i],0), sum_xx=hist_ext.reduce((a,b)=>a+bb,0);\n let slope=(nsum_xy-sum_xsum_y)/(nsum_xx-sum_xsum_x);\n let inter=(sum_y-slopesum_x)/n;\n \n let t_pred=t_ext-0.5; // Prediccion conservadora\n let timp_ia=Math.max(30,Math.min(55,slopet_pred+inter));\n let timp_fija=35+(20)*(15-t_pred)/(17); // Curva fija tipica\n \n msg.payload={\n hora:new Date().toLocaleTimeString('es'),\n t_ext:t_ext.toFixed(1)+' C', t_pred:t_pred.toFixed(1)+' C',\n timp_real:t_imp.toFixed(1)+' C', timp_ia:timp_ia.toFixed(1)+' C',\n timp_fija:timp_fija.toFixed(1)+' C', cop:cop.toFixed(2),\n entrenados:n, accion:timp_ia>t_imp+1?'TOP29++':timp_ia<t_imp-1?'TOP29--':'OK'\n };\n}else{\n msg.payload={error:Entrenando IA... ${Math.min(20,20-n)}/20 datos};\n}\nreturn msg;","outputs":1,"x":1080,"y":130,"wires":[["dashboard","modbus"]]},
{"id":"dashboard","type":"debug","z":"aerotermia_ia","name":"LIVE (20s)","active":true,"x":1300,"y":100,"wires":[]},
{"id":"status","type":"ui_text","z":"aerotermia_ia","group":"dashboard","order":1,"width":6,"format":"{{msg.payload}}","x":650,"y":60,"wires":[]},
{"id":"gauge_ia","type":"ui_gauge","z":"aerotermia_ia","group":"dashboard","order":2,"width":6,"title":"T_imp IA","label":"C","format":"{{value | number:1}}","min":25,"max":60,"colors":["#00b500","#e6e600","#ca3838"],"x":1300,"y":140,"wires":[["payload.timp_ia"]]},
{"id":"table_live","type":"ui_table","z":"aerotermia_ia","group":"dashboard","order":3,"width":12,"height":6,"columns":[{"field":"hora","title":"Hora","width":"80px"},{"field":"t_ext","title":"Ext","width":"70px"},{"field":"timp_real","title":"Real","width":"70px"},{"field":"timp_ia","title":"IA","width":"70px"},{"field":"timp_fija","title":"Fija","width":"70px"},{"field":"cop","title":"COP","width":"60px"},{"field":"accion","title":"Accion","width":"100px"}],"x":1300,"y":190,"wires":[]},
{"id":"dashboard_group","type":"ui_group","name":"Aerotermia IA","tab":"ui_tab","order":1,"width":12},
{"id":"ui_tab","type":"ui_tab","name":"Dashboard IA","icon":"dashboard","order":1},
{"id":"heishamon_broker","type":"mqtt-broker","name":"HeishaMon","broker":"192.168.1.XX","port":"1883","clientid":"aquarea_ia","autoConnect":true,"usetls":false}
]

AI IMPLEMENTATION: ADAPTIVE LINEAR REGRESSION MODEL

This model automatically "learns" the optimal curve of your specific heat pump by analyzing real HeishaMon data (Outdoor Temp TOP27 vs. Flow Temp TOP6 and COP TOP114) over 24h+.

AI HEAT PUMP + AEMET + HEISHAMON NODE-RED: FULL EXPLANATION
(Optimized for Panasonic Aquarea T-CAP - 15-20% COP Gain)

WHAT THE AI DOES
It replaces the generic Panasonic fixed curve (TOP29-32) with an INTELLIGENT curve that learns from YOUR specific home insulation and system response + AEMET weather forecasting.

FIXED Curve (Standard): Outdoor 15C -> 50C | -2C -> 55C

AI Curve (Learned): Outdoor 15C -> 48C | -2C -> 52C (YOUR HOME)

RESULT: +15% COP -> Approx. 0.50e to 1.00e daily savings.

AI ALGORITHM STEP-BY-STEP

DATA COLLECTION (Every 20s)

AEMET API: Real-time outdoor temp forecast (via local INE code).

HEISHAMON: TOP6 (Flow Temp), TOP27 (Outdoor Temp), TOP114 (Actual COP).

HISTORY: 100 continuous data points -> "AI MEMORY".

AUTOMATIC TRAINING
The AI calculates: Optimal_Flow = (slope * Outdoor_Temp) + intercept
Formula: slope = (n*sum(xy) - sum(x)sum(y)) / (nsum(x^2) - (sum(x))^2)
(This generates YOUR REAL efficiency curve).

PREDICTION (Every 20s)

AEMET Predicts: 4.3C for the next hour.
AI Calculates: Flow_Temp_AI = (-1.18 * 4.3) + 57.2 = 48.1C
Standard Panasonic: 50.5C
ACTION: Lower TOP29 by 2.4C -> Immediate +5% COP gain.

QUICK COMPARISON

PERSONALIZATION:

Fixed: Generic
AI+AEMET: YOUR HOME
GAIN: +10-15% COP

PREDICTION:

Fixed: Reactive
AI+AEMET: Proactive (Anticipates weather)
GAIN: -20% On/Off Cycles

VALIDATION:

Fixed: Blind logic
AI+AEMET: Real-time COP monitoring
GAIN: Maximum Efficiency

ADAPTABILITY:

Fixed: Static
AI+AEMET: Self-improving monthly
GAIN: Constant +5% improvement

Grok or Perplexity act as the 'supervisor' that analyzes long-term trends or helps optimize the code during extreme weather conditions that simple mathematics cannot explain

https://opendata.aemet.es/centrodedescargas/inicio

From now on, I'm asking you, PLEASE, to close my account, to ban me from this forum because I don't want to continue arguing or proving anything, and as I mentioned before, I respect your comment.
From now on, I will no longer answer your questions.
Thank you for your patience and for wasting your valuable time reading my comments