diff --git a/custom_components/weishaupt_modbus/kennfeld.py b/custom_components/weishaupt_modbus/kennfeld.py
new file mode 100644
index 0000000..4d660eb
--- /dev/null
+++ b/custom_components/weishaupt_modbus/kennfeld.py
@@ -0,0 +1,71 @@
+from scipy.interpolate import CubicSpline
+import numpy as np
+
+# visual debugging ;-)
+# import matplotlib.pyplot as plt
+
+# these are values extracted from the characteristic curves of heating power found ion the documentation of my heat pump.
+# there are two diagrams: 
+#  - heating power vs. outside temperature @ 35 °C flow temperature
+#  - heating power vs. outside temperature @ 55 °C flow temperature
+known_x = [   -30,   -25,   -22,   -20,   -15,   -10,    -5,     0,     5,    10,    15,    20,    25,    30,    35,    40]
+# known power values read out from the graphs plotted in documentation. Only 35 °C and 55 °C available
+known_y = [[ 5700,  5700,  5700,  5700,  6290,  7580,  8660,  9625, 10300, 10580, 10750, 10790, 10830, 11000, 11000, 11000 ],
+           [ 5700,  5700,  5700,  5700,  6860,  7300,  8150,  9500, 10300, 10580, 10750, 10790, 10830, 11000, 11000, 11000 ]]
+
+# the known x values for linear interpolation
+known_t = [35, 55]
+
+# the aim is generating a 2D power map that gives back the actual power for a certain flow temperature and a given outside temperature
+# the map should have values on every integer temperature point
+# at first, all flow temoperatures are lineary interpolated 
+steps = 21
+r_to_interpolate = np.linspace(35, 55, steps)
+
+# build the matrix with linear interpolated samples
+# 1st and last row are populated by known values from diagrem, the rest is zero
+interp_y = []
+interp_y.append(known_y[0])
+v = np.linspace(0, steps-3, steps-2)
+for idx in v:
+    interp_y.append(np.zeros_like(known_x))
+interp_y.append(known_y[1])
+
+# visual debugging ;-)
+#plt.plot(np.transpose(interp_y))
+#plt.ylabel('Max Power')
+#plt.xlabel('°C')
+#plt.show()
+
+for idx in range(0, len(known_x)):
+    # the known y for every column
+    yk = [interp_y[0][idx], interp_y[steps-1][idx]]
+    
+    #linear interpolation
+    ip = np.interp(r_to_interpolate, known_t, yk)
+    
+    # sort the interpolated values into the array
+    for r in range(0, len(r_to_interpolate)):
+        interp_y[r][idx] = ip[r]
+
+# visual debugging ;-)
+#plt.plot(np.transpose(interp_y))
+#plt.ylabel('Max Power')
+#plt.xlabel('°C')
+#plt.show()
+
+# at second step, power vs. outside temp are interpolated using cubic splines
+# the output matrix
+max_power = []
+# we want to have samples at every integer °C
+t = np.linspace(-30, 40, 71)
+# cubic spline interpolation of power curves
+for idx in range(0, len(r_to_interpolate)):
+    f = CubicSpline(known_x, interp_y[idx], bc_type='natural')
+    max_power.append(f(t))
+
+# visual debugging ;-)
+#plt.plot(t,np.transpose(max_power))
+#plt.ylabel('Max Power')
+#plt.xlabel('°C')
+#plt.show()